Merge remote-tracking branch 'origin/codex/platform-cost-tracking' into combined-preview-test

1. Rewrite tautological env_test.py TestClaudeCodeTmpdir tests to call
   build_sdk_env(sdk_cwd=...) directly instead of copy-pasting the
   if-sdk_cwd pattern. Moved CLAUDE_CODE_TMPDIR logic into build_sdk_env().
2. Add DEL (\x7f), C1 (\x80-\x9f), BiDi, and zero-width chars to
   security_hooks_test.py sanitization test inputs.
3. Promote _sanitize() from closure to module-level pure function.
4. Fix GenericTool.tsx "model may poll again" -> user-friendly message.
5. Replace `as never` with @ts-expect-error + comment in useChatSession.ts.
6. Extract "Agent"/"Task"/"TaskOutput" string literals to named constants
   in helpers.ts, imported in GenericTool.tsx.
7. Extend _sanitize() to strip Unicode BiDi overrides (U+202A-U+202E,
   U+2066-U+2069) and zero-width characters (U+200B-U+200F, U+FEFF).
8. Document background agent slot lifecycle limitation in security_hooks.py
   (SubagentStop doesn't fire reliably for background agents).

.claude/settings.json

@@ -1,10 +0,0 @@
-{
-  "permissions": {
-    "allowedTools": [
-      "Read", "Grep", "Glob",
-      "Bash(ls:*)", "Bash(cat:*)", "Bash(grep:*)", "Bash(find:*)",
-      "Bash(git status:*)", "Bash(git diff:*)", "Bash(git log:*)", "Bash(git worktree:*)",
-      "Bash(tmux:*)", "Bash(sleep:*)", "Bash(branchlet:*)"
-    ]
-  }
-}

.claude/skills/orchestrate/SKILL.md

@@ -1,545 +0,0 @@
----
-name: orchestrate
-description: "Meta-agent supervisor that manages a fleet of Claude Code agents running in tmux windows. Auto-discovers spare worktrees, spawns agents, monitors state, kicks idle agents, approves safe confirmations, and recycles worktrees when done. TRIGGER when user asks to supervise agents, run parallel tasks, manage worktrees, check agent status, or orchestrate parallel work."
-user-invocable: true
-argument-hint: "any free text — e.g. 'start 3 agents on X Y Z', 'show status', 'add task: implement feature A', 'stop', 'how many are free?'"
-metadata:
-  author: autogpt-team
-  version: "6.0.0"
----
-
-# Orchestrate — Agent Fleet Supervisor
-
-One tmux session, N windows — each window is one agent working in its own worktree. Speak naturally; Claude maps your intent to the right scripts.
-
-## Scripts
-
-```bash
-SKILLS_DIR=$(git rev-parse --show-toplevel)/.claude/skills/orchestrate/scripts
-STATE_FILE=~/.claude/orchestrator-state.json
-```
-
-| Script | Purpose |
-|---|---|
-| `find-spare.sh [REPO_ROOT]` | List free worktrees — one `PATH BRANCH` per line |
-| `spawn-agent.sh SESSION PATH SPARE NEW_BRANCH OBJECTIVE [PR_NUMBER] [STEPS...]` | Create window + checkout branch + launch claude + send task. **Stdout: `SESSION:WIN` only** |
-| `recycle-agent.sh WINDOW PATH SPARE_BRANCH` | Kill window + restore spare branch |
-| `run-loop.sh` | **Mechanical babysitter** — idle restart + dialog approval + recycle on ORCHESTRATOR:DONE + supervisor health check + all-done notification |
-| `verify-complete.sh WINDOW` | Verify PR is done: checkpoints ✓ + 0 unresolved threads + CI green + no fresh CHANGES_REQUESTED. Repo auto-derived from state file `.repo` or git remote. |
-| `notify.sh MESSAGE` | Send notification via Discord webhook (env `DISCORD_WEBHOOK_URL` or state `.discord_webhook`), macOS notification center, and stdout |
-| `capacity.sh [REPO_ROOT]` | Print available + in-use worktrees |
-| `status.sh` | Print fleet status + live pane commands |
-| `poll-cycle.sh` | One monitoring cycle — classifies panes, tracks checkpoints, returns JSON action array |
-| `classify-pane.sh WINDOW` | Classify one pane state |
-
-## Supervision model
-
-```
-Orchestrating Claude (this Claude session — IS the supervisor)
-  └── Reads pane output, checks CI, intervenes with targeted guidance
-        run-loop.sh (separate tmux window, every 30s)
-          └── Mechanical only: idle restart, dialog approval, recycle on ORCHESTRATOR:DONE
-```
-
-**You (the orchestrating Claude)** are the supervisor. After spawning agents, stay in this conversation and actively monitor: poll each agent's pane every 2-3 minutes, check CI, nudge stalled agents, and verify completions. Do not spawn a separate supervisor Claude window — it loses context, is hard to observe, and compounds context compression problems.
-
-**run-loop.sh** is the mechanical layer — zero tokens, handles things that need no judgment: restart crashed agents, press Enter on dialogs, recycle completed worktrees (only after `verify-complete.sh` passes).
-
-## Checkpoint protocol
-
-Agents output checkpoints as they complete each required step:
-
-```
-CHECKPOINT:<step-name>
-```
-
-Required steps are passed as args to `spawn-agent.sh` (e.g. `pr-address pr-test`). `run-loop.sh` will not recycle a window until all required checkpoints are found in the pane output. If `verify-complete.sh` fails, the agent is re-briefed automatically.
-
-## Worktree lifecycle
-
-```text
-spare/N branch  →  spawn-agent.sh (--session-id UUID)  →  window + feat/branch + claude running
-                                                                 ↓
-                                               CHECKPOINT:<step> (as steps complete)
-                                                                 ↓
-                                                        ORCHESTRATOR:DONE
-                                                                 ↓
-                                    verify-complete.sh: checkpoints ✓ + 0 threads + CI green + no fresh CHANGES_REQUESTED
-                                                                 ↓
-                                              state → "done", notify, window KEPT OPEN
-                                                                 ↓
-                              user/orchestrator explicitly requests recycle
-                                                                 ↓
-                                         recycle-agent.sh → spare/N (free again)
-```
-
-**Windows are never auto-killed.** The worktree stays on its branch, the session stays alive. The agent is done working but the window, git state, and Claude session are all preserved until you choose to recycle.
-
-**To resume a done or crashed session:**
-```bash
-# Resume by stored session ID (preferred — exact session, full context)
-claude --resume SESSION_ID --permission-mode bypassPermissions
-
-# Or resume most recent session in that worktree directory
-cd /path/to/worktree && claude --continue --permission-mode bypassPermissions
-```
-
-**To manually recycle when ready:**
-```bash
-bash ~/.claude/orchestrator/scripts/recycle-agent.sh SESSION:WIN WORKTREE_PATH spare/N
-# Then update state:
-jq --arg w "SESSION:WIN" '.agents |= map(if .window == $w then .state = "recycled" else . end)' \
-  ~/.claude/orchestrator-state.json > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-```
-
-## State file (`~/.claude/orchestrator-state.json`)
-
-Never committed to git. You maintain this file directly using `jq` + atomic writes (`.tmp` → `mv`).
-
-```json
-{
-  "active": true,
-  "tmux_session": "autogpt1",
-  "idle_threshold_seconds": 300,
-  "loop_window": "autogpt1:5",
-  "repo": "Significant-Gravitas/AutoGPT",
-  "discord_webhook": "https://discord.com/api/webhooks/...",
-  "last_poll_at": 0,
-  "agents": [
-    {
-      "window": "autogpt1:3",
-      "worktree": "AutoGPT6",
-      "worktree_path": "/path/to/AutoGPT6",
-      "spare_branch": "spare/6",
-      "branch": "feat/my-feature",
-      "objective": "Implement X and open a PR",
-      "pr_number": "12345",
-      "session_id": "550e8400-e29b-41d4-a716-446655440000",
-      "steps": ["pr-address", "pr-test"],
-      "checkpoints": ["pr-address"],
-      "state": "running",
-      "last_output_hash": "",
-      "last_seen_at": 0,
-      "spawned_at": 0,
-      "idle_since": 0,
-      "revision_count": 0,
-      "last_rebriefed_at": 0
-    }
-  ]
-}
-```
-
-Top-level optional fields:
-- `repo` — GitHub `owner/repo` for CI/thread checks. Auto-derived from git remote if omitted.
-- `discord_webhook` — Discord webhook URL for completion notifications. Also reads `DISCORD_WEBHOOK_URL` env var.
-
-Per-agent fields:
-- `session_id` — UUID passed to `claude --session-id` at spawn; use with `claude --resume UUID` to restore exact session context after a crash or window close.
-- `last_rebriefed_at` — Unix timestamp of last re-brief; enforces 5-min cooldown to prevent spam.
-
-Agent states: `running` | `idle` | `stuck` | `waiting_approval` | `complete` | `done` | `escalated`
-
-`done` means verified complete — window is still open, session still alive, worktree still on task branch. Not recycled yet.
-
-## Serial /pr-test rule
-
-`/pr-test` and `/pr-test --fix` run local Docker + integration tests that use shared ports, a shared database, and shared build caches. **Running two `/pr-test` jobs simultaneously will cause port conflicts and database corruption.**
-
-**Rule: only one `/pr-test` runs at a time. The orchestrator serializes them.**
-
-You (the orchestrating Claude) own the test queue:
-1. Agents do `pr-review` and `pr-address` in parallel — that's safe (they only push code and reply to GitHub).
-2. When a PR needs local testing, add it to your mental queue — don't give agents a `pr-test` step.
-3. Run `/pr-test https://github.com/OWNER/REPO/pull/PR_NUMBER --fix` yourself, sequentially.
-4. Feed results back to the relevant agent via `tmux send-keys`:
-   ```bash
-   tmux send-keys -t SESSION:WIN "Local tests for PR #N: <paste failure output or 'all passed'>. Fix any failures and push, then output ORCHESTRATOR:DONE."
-   sleep 0.3
-   tmux send-keys -t SESSION:WIN Enter
-   ```
-5. Wait for CI to confirm green before marking the agent done.
-
-If multiple PRs need testing at the same time, pick the one furthest along (fewest pending CI checks) and test it first. Only start the next test after the previous one completes.
-
-## Session restore (tested and confirmed)
-
-Agent sessions are saved to disk. To restore a closed or crashed session:
-
-```bash
-# If session_id is in state (preferred):
-NEW_WIN=$(tmux new-window -t SESSION -n WORKTREE_NAME -P -F '#{window_index}')
-tmux send-keys -t "SESSION:${NEW_WIN}" "cd /path/to/worktree && claude --resume SESSION_ID --permission-mode bypassPermissions" Enter
-
-# If no session_id (use --continue for most recent session in that directory):
-tmux send-keys -t "SESSION:${NEW_WIN}" "cd /path/to/worktree && claude --continue --permission-mode bypassPermissions" Enter
-```
-
-`--continue` restores the full conversation history including all tool calls, file edits, and context. The agent resumes exactly where it left off. After restoring, update the window address in the state file:
-
-```bash
-jq --arg old "SESSION:OLD_WIN" --arg new "SESSION:NEW_WIN" \
-  '(.agents[] | select(.window == $old)).window = $new' \
-  ~/.claude/orchestrator-state.json > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-```
-
-## Intent → action mapping
-
-Match the user's message to one of these intents:
-
-| The user says something like… | What to do |
-|---|---|
-| "status", "what's running", "show agents" | Run `status.sh` + `capacity.sh`, show output |
-| "how many free", "capacity", "available worktrees" | Run `capacity.sh`, show output |
-| "start N agents on X, Y, Z" or "run these tasks: …" | See **Spawning agents** below |
-| "add task: …", "add one more agent for …" | See **Adding an agent** below |
-| "stop", "shut down", "pause the fleet" | See **Stopping** below |
-| "poll", "check now", "run a cycle" | Run `poll-cycle.sh`, process actions |
-| "recycle window X", "free up autogpt3" | Run `recycle-agent.sh` directly |
-
-When the intent is ambiguous, show capacity first and ask what tasks to run.
-
-## Spawning agents
-
-### 1. Resolve tmux session
-
-```bash
-tmux list-sessions -F "#{session_name}: #{session_windows} windows" 2>/dev/null
-```
-
-Use an existing session. **Never create a tmux session from within Claude** — it becomes a child of Claude's process and dies when the session ends. If no session exists, tell the user to run `tmux new-session -d -s autogpt1` in their terminal first, then re-invoke `/orchestrate`.
-
-### 2. Show available capacity
-
-```bash
-bash $SKILLS_DIR/capacity.sh $(git rev-parse --show-toplevel)
-```
-
-### 3. Collect tasks from the user
-
-For each task, gather:
-- **objective** — what to do (e.g. "implement feature X and open a PR")
-- **branch name** — e.g. `feat/my-feature` (derive from objective if not given)
-- **pr_number** — GitHub PR number if working on an existing PR (for verification)
-- **steps** — required checkpoint names in order (e.g. `pr-address pr-test`) — derive from objective
-
-Ask for `idle_threshold_seconds` only if the user mentions it (default: 300).
-
-Never ask the user to specify a worktree — auto-assign from `find-spare.sh`.
-
-### 4. Spawn one agent per task
-
-```bash
-# Get ordered list of spare worktrees
-SPARE_LIST=$(bash $SKILLS_DIR/find-spare.sh $(git rev-parse --show-toplevel))
-
-# For each task, take the next spare line:
-WORKTREE_PATH=$(echo "$SPARE_LINE" | awk '{print $1}')
-SPARE_BRANCH=$(echo "$SPARE_LINE" | awk '{print $2}')
-
-# With PR number and required steps:
-WINDOW=$(bash $SKILLS_DIR/spawn-agent.sh "$SESSION" "$WORKTREE_PATH" "$SPARE_BRANCH" "$NEW_BRANCH" "$OBJECTIVE" "$PR_NUMBER" "pr-address" "pr-test")
-
-# Without PR (new work):
-WINDOW=$(bash $SKILLS_DIR/spawn-agent.sh "$SESSION" "$WORKTREE_PATH" "$SPARE_BRANCH" "$NEW_BRANCH" "$OBJECTIVE")
-```
-
-Build an agent record and append it to the state file. If the state file doesn't exist yet, initialize it:
-
-```bash
-# Derive repo from git remote (used by verify-complete.sh + supervisor)
-REPO=$(git remote get-url origin 2>/dev/null | sed 's|.*github\.com[:/]||; s|\.git$||' || echo "")
-
-jq -n \
-  --arg session "$SESSION" \
-  --arg repo "$REPO" \
-  --argjson threshold 300 \
-  '{active:true, tmux_session:$session, idle_threshold_seconds:$threshold,
-    repo:$repo, loop_window:null, supervisor_window:null, last_poll_at:0, agents:[]}' \
-  > ~/.claude/orchestrator-state.json
-```
-
-Optionally add a Discord webhook for completion notifications:
-```bash
-jq --arg hook "$DISCORD_WEBHOOK_URL" '.discord_webhook = $hook' ~/.claude/orchestrator-state.json \
-  > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-```
-
-`spawn-agent.sh` writes the initial agent record (window, worktree_path, branch, objective, state, etc.) to the state file automatically — **do not append the record again after calling it.** The record already exists and `pr_number`/`steps` are patched in by the script itself.
-
-### 5. Start the mechanical babysitter
-
-```bash
-LOOP_WIN=$(tmux new-window -t "$SESSION" -n "orchestrator" -P -F '#{window_index}')
-LOOP_WINDOW="${SESSION}:${LOOP_WIN}"
-tmux send-keys -t "$LOOP_WINDOW" "bash $SKILLS_DIR/run-loop.sh" Enter
-
-jq --arg w "$LOOP_WINDOW" '.loop_window = $w' ~/.claude/orchestrator-state.json \
-  > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-```
-
-### 6. Begin supervising directly in this conversation
-
-You are the supervisor. After spawning, immediately start your first poll loop (see **Supervisor duties** below) and continue every 2-3 minutes. Do NOT spawn a separate supervisor Claude window.
-
-## Adding an agent
-
-Find the next spare worktree, then spawn and append to state — same as steps 2–4 above but for a single task. If no spare worktrees are available, tell the user.
-
-## Supervisor duties (YOUR job, every 2-3 min in this conversation)
-
-You are the supervisor. Run this poll loop directly in your Claude session — not in a separate window.
-
-### Poll loop mechanism
-
-You are reactive — you only act when a tool completes or the user sends a message. To create a self-sustaining poll loop without user involvement:
-
-1. Start each poll with `run_in_background: true` + a sleep before the work:
-   ```bash
-   sleep 120 && tmux capture-pane -t autogpt1:0 -p -S -200 | tail -40
-   # + similar for each active window
-   ```
-2. When the background job notifies you, read the pane output and take action.
-3. Immediately schedule the next background poll — this keeps the loop alive.
-4. Stop scheduling when all agents are done/escalated.
-
-**Never tell the user "I'll poll every 2-3 minutes"** — that does nothing without a trigger. Start the background job instead.
-
-### Each poll: what to check
-
-```bash
-# 1. Read state
-cat ~/.claude/orchestrator-state.json | jq '.agents[] | {window, worktree, branch, state, pr_number, checkpoints}'
-
-# 2. For each running/stuck/idle agent, capture pane
-tmux capture-pane -t SESSION:WIN -p -S -200 | tail -60
-```
-
-For each agent, decide:
-
-| What you see | Action |
-|---|---|
-| Spinner / tools running | Do nothing — agent is working |
-| Idle `❯` prompt, no `ORCHESTRATOR:DONE` | Stalled — send specific nudge with objective from state |
-| Stuck in error loop | Send targeted fix with exact error + solution |
-| Waiting for input / question | Answer and unblock via `tmux send-keys` |
-| CI red | `gh pr checks PR_NUMBER --repo REPO` → tell agent exactly what's failing |
-| Context compacted / agent lost | Send recovery: `cat ~/.claude/orchestrator-state.json | jq '.agents[] | select(.window=="WIN")'` + `gh pr view PR_NUMBER --json title,body` |
-| `ORCHESTRATOR:DONE` in output | Run `verify-complete.sh` — if it fails, re-brief with specific reason |
-
-### Strict ORCHESTRATOR:DONE gate
-
-`verify-complete.sh` handles the main checks automatically (checkpoints, threads, CI green, spawned_at, and CHANGES_REQUESTED). Run it:
-
-**CHANGES_REQUESTED staleness rule**: a `CHANGES_REQUESTED` review only blocks if it was submitted *after* the latest commit. If the latest commit postdates the review, the review is considered stale (feedback already addressed) and does not block. This avoids false negatives when a bot reviewer hasn't re-reviewed after the agent's fixing commits.
-
-```bash
-SKILLS_DIR=~/.claude/orchestrator/scripts
-bash $SKILLS_DIR/verify-complete.sh SESSION:WIN
-```
-
-If it passes → run-loop.sh will recycle the window automatically. No manual action needed.
-If it fails → re-brief the agent with the failure reason. Never manually mark state `done` to bypass this.
-
-### Re-brief a stalled agent
-
-```bash
-OBJ=$(jq -r --arg w SESSION:WIN '.agents[] | select(.window==$w) | .objective' ~/.claude/orchestrator-state.json)
-PR=$(jq -r --arg w SESSION:WIN '.agents[] | select(.window==$w) | .pr_number' ~/.claude/orchestrator-state.json)
-tmux send-keys -t SESSION:WIN "You appear stalled. Your objective: $OBJ. Check: gh pr view $PR --json title,body,headRefName to reorient."
-sleep 0.3
-tmux send-keys -t SESSION:WIN Enter
-```
-
-If `image_path` is set on the agent record, include: "Re-read context at IMAGE_PATH with the Read tool."
-
-## Self-recovery protocol (agents)
-
-spawn-agent.sh automatically includes this instruction in every objective:
-
-> If your context compacts and you lose track of what to do, run:
-> `cat ~/.claude/orchestrator-state.json | jq '.agents[] | select(.window=="SESSION:WIN")'`
-> and `gh pr view PR_NUMBER --json title,body,headRefName` to reorient.
-> Output each completed step as `CHECKPOINT:<step-name>` on its own line.
-
-## Passing images and screenshots to agents
-
-`tmux send-keys` is text-only — you cannot paste a raw image into a pane. To give an agent visual context (screenshots, diagrams, mockups):
-
-1. **Save the image to a temp file** with a stable path:
-   ```bash
-   # If the user drags in a screenshot or you receive a file path:
-   IMAGE_PATH="/tmp/orchestrator-context-$(date +%s).png"
-   cp "$USER_PROVIDED_PATH" "$IMAGE_PATH"
-   ```
-
-2. **Reference the path in the objective string**:
-   ```bash
-   OBJECTIVE="Implement the layout shown in /tmp/orchestrator-context-1234567890.png. Read that image first with the Read tool to understand the design."
-   ```
-
-3. The agent uses its `Read` tool to view the image at startup — Claude Code agents are multimodal and can read image files directly.
-
-**Rule**: always use `/tmp/orchestrator-context-<timestamp>.png` as the naming convention so the supervisor knows what to look for if it needs to re-brief an agent with the same image.
-
----
-
-## Orchestrator final evaluation (YOU decide, not the script)
-
-`verify-complete.sh` is a gate — it blocks premature marking. But it cannot tell you if the work is actually good. That is YOUR job.
-
-When run-loop marks an agent `pending_evaluation` and you're notified, do all of these before marking done:
-
-### 1. Run /pr-test (required, serialized, use TodoWrite to queue)
-
-`/pr-test` is the only reliable confirmation that the objective is actually met. Run it yourself, not the agent.
-
-**When multiple PRs reach `pending_evaluation` at the same time, use TodoWrite to queue them:**
-```
-- [ ] /pr-test PR #12636 — fix copilot retry logic
-- [ ] /pr-test PR #12699 — builder chat panel
-```
-Run one at a time. Check off as you go.
-
-```
-/pr-test https://github.com/Significant-Gravitas/AutoGPT/pull/PR_NUMBER
-```
-
-**/pr-test can be lazy** — if it gives vague output, re-run with full context:
-
-```
-/pr-test https://github.com/OWNER/REPO/pull/PR_NUMBER
-Context: This PR implements <objective from state file>. Key files: <list>.
-Please verify: <specific behaviors to check>.
-```
-
-Only one `/pr-test` at a time — they share ports and DB.
-
-### /pr-test result evaluation
-
-**PARTIAL on any headline feature scenario is an immediate blocker.** Do not approve, do not mark done, do not let the agent output `ORCHESTRATOR:DONE`.
-
-| `/pr-test` result | Action |
-|---|---|
-| All headline scenarios **PASS** | Proceed to evaluation step 2 |
-| Any headline scenario **PARTIAL** | Re-brief the agent immediately — see below |
-| Any headline scenario **FAIL** | Re-brief the agent immediately |
-
-**What PARTIAL means**: the feature is only partly working. Example: the Apply button never appeared, or the AI returned no action blocks. The agent addressed part of the objective but not all of it.
-
-**When any headline scenario is PARTIAL or FAIL:**
-
-1. Do NOT mark the agent done or accept `ORCHESTRATOR:DONE`
-2. Re-brief the agent with the specific scenario that failed and what was missing:
-   ```bash
-   tmux send-keys -t SESSION:WIN "PARTIAL result on /pr-test — S5 (Apply button) never appeared. The AI must output JSON action blocks for the Apply button to render. Fix this before re-running /pr-test."
-   sleep 0.3
-   tmux send-keys -t SESSION:WIN Enter
-   ```
-3. Set state back to `running`:
-   ```bash
-   jq --arg w "SESSION:WIN" '(.agents[] | select(.window == $w)).state = "running"' \
-     ~/.claude/orchestrator-state.json > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-   ```
-4. Wait for new `ORCHESTRATOR:DONE`, then re-run `/pr-test` from scratch
-
-**Rule: only ALL-PASS qualifies for approval.** A mix of PASS + PARTIAL is a failure.
-
-> **Why this matters**: PR #12699 was wrongly approved with S5 PARTIAL — the AI never output JSON action blocks so the Apply button never appeared. The fix was already in the agent's reach but slipped through because PARTIAL was not treated as blocking.
-
-### 2. Do your own evaluation
-
-1. **Read the PR diff and objective** — does the code actually implement what was asked? Is anything obviously missing or half-done?
-2. **Read the resolved threads** — were comments addressed with real fixes, or just dismissed/resolved without changes?
-3. **Check CI run names** — any suspicious retries that shouldn't have passed?
-4. **Check the PR description** — title, summary, test plan complete?
-
-### 3. Decide
-
-- `/pr-test` all scenarios PASS + evaluation looks good → mark `done` in state, tell the user the PR is ready, ask if window should be closed
-- `/pr-test` any scenario PARTIAL or FAIL → re-brief the agent with the specific failing scenario, set state back to `running` (see `/pr-test result evaluation` above)
-- Evaluation finds gaps even with all PASS → re-brief the agent with specific gaps, set state back to `running`
-
-**Never mark done based purely on script output.** You hold the full objective context; the script does not.
-
-```bash
-# Mark done after your positive evaluation:
-jq --arg w "SESSION:WIN" '(.agents[] | select(.window == $w)).state = "done"' \
-  ~/.claude/orchestrator-state.json > /tmp/orch.tmp && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-```
-
-## When to stop the fleet
-
-Stop the fleet (`active = false`) when **all** of the following are true:
-
-| Check | How to verify |
-|---|---|
-| All agents are `done` or `escalated` | `jq '[.agents[] | select(.state | test("running\|stuck\|idle\|waiting_approval"))] | length' ~/.claude/orchestrator-state.json` == 0 |
-| All PRs have 0 unresolved review threads | GraphQL `isResolved` check per PR |
-| All PRs have green CI **on a run triggered after the agent's last push** | `gh run list --branch BRANCH --limit 1` timestamp > `spawned_at` in state |
-| No fresh CHANGES_REQUESTED (after latest commit) | `verify-complete.sh` checks this — stale pre-commit reviews are ignored |
-| No agents are `escalated` without human review | If any are escalated, surface to user first |
-
-**Do NOT stop just because agents output `ORCHESTRATOR:DONE`.** That is a signal to verify, not a signal to stop.
-
-**Do stop** if the user explicitly says "stop", "shut down", or "kill everything", even with agents still running.
-
-```bash
-# Graceful stop
-jq '.active = false' ~/.claude/orchestrator-state.json > /tmp/orch.tmp \
-  && mv /tmp/orch.tmp ~/.claude/orchestrator-state.json
-
-LOOP_WINDOW=$(jq -r '.loop_window // ""' ~/.claude/orchestrator-state.json)
-[ -n "$LOOP_WINDOW" ] && tmux kill-window -t "$LOOP_WINDOW" 2>/dev/null || true
-```
-
-Does **not** recycle running worktrees — agents may still be mid-task. Run `capacity.sh` to see what's still in progress.
-
-## tmux send-keys pattern
-
-**Always split long messages into text + Enter as two separate calls with a sleep between them.** If sent as one call (`"text" Enter`), Enter can fire before the full string is buffered into Claude's input — leaving the message stuck as `[Pasted text +N lines]` unsent.
-
-```bash
-# CORRECT — text then Enter separately
-tmux send-keys -t "$WINDOW" "your long message here"
-sleep 0.3
-tmux send-keys -t "$WINDOW" Enter
-
-# WRONG — Enter may fire before text is buffered
-tmux send-keys -t "$WINDOW" "your long message here" Enter
-```
-
-Short single-character sends (`y`, `Down`, empty Enter for dialog approval) are safe to combine since they have no buffering lag.
-
----
-
-## Protected worktrees
-
-Some worktrees must **never** be used as spare worktrees for agent tasks because they host files critical to the orchestrator itself:
-
-| Worktree | Protected branch | Why |
-|---|---|---|
-| `AutoGPT1` | `dx/orchestrate-skill` | Hosts the orchestrate skill scripts. `recycle-agent.sh` would check out `spare/1`, wiping `.claude/skills/` and breaking all subsequent `spawn-agent.sh` calls. |
-
-**Rule**: when selecting spare worktrees via `find-spare.sh`, skip any worktree whose CURRENT branch matches a protected branch. If you accidentally spawn an agent in a protected worktree, do not let `recycle-agent.sh` run on it — manually restore the branch after the agent finishes.
-
-When `dx/orchestrate-skill` is merged into `dev`, `AutoGPT1` becomes a normal spare again.
-
----
-
-## Key rules
-
-1. **Scripts do all the heavy lifting** — don't reimplement their logic inline in this file
-2. **Never ask the user to pick a worktree** — auto-assign from `find-spare.sh` output
-3. **Never restart a running agent** — only restart on `idle` kicks (foreground is a shell)
-4. **Auto-dismiss settings dialogs** — if "Enter to confirm" appears, send Down+Enter
-5. **Always `--permission-mode bypassPermissions`** on every spawn
-6. **Escalate after 3 kicks** — mark `escalated`, surface to user
-7. **Atomic state writes** — always write to `.tmp` then `mv`
-8. **Never approve destructive commands** outside the worktree scope — when in doubt, escalate
-9. **Never recycle without verification** — `verify-complete.sh` must pass before recycling
-10. **No TASK.md files** — commit risk; use state file + `gh pr view` for agent context persistence
-11. **Re-brief stalled agents** — read objective from state file + `gh pr view`, send via tmux
-12. **ORCHESTRATOR:DONE is a signal to verify, not to accept** — always run `verify-complete.sh` and check CI run timestamp before recycling
-13. **Protected worktrees** — never use the worktree hosting the skill scripts as a spare
-14. **Images via file path** — save screenshots to `/tmp/orchestrator-context-<ts>.png`, pass path in objective; agents read with the `Read` tool
-15. **Split send-keys** — always separate text and Enter with `sleep 0.3` between calls for long strings

.claude/skills/orchestrate/scripts/capacity.sh

@@ -1,43 +0,0 @@
-#!/usr/bin/env bash
-# capacity.sh — show fleet capacity: available spare worktrees + in-use agents
-#
-# Usage: capacity.sh [REPO_ROOT]
-#   REPO_ROOT defaults to the root worktree of the current git repo.
-#
-# Reads: ~/.claude/orchestrator-state.json (skipped if missing or corrupt)
-
-set -euo pipefail
-
-SCRIPTS_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-REPO_ROOT="${1:-$(git rev-parse --show-toplevel 2>/dev/null || echo "")}"
-
-echo "=== Available (spare) worktrees ==="
-if [ -n "$REPO_ROOT" ]; then
-  SPARE=$("$SCRIPTS_DIR/find-spare.sh" "$REPO_ROOT" 2>/dev/null || echo "")
-else
-  SPARE=$("$SCRIPTS_DIR/find-spare.sh" 2>/dev/null || echo "")
-fi
-
-if [ -z "$SPARE" ]; then
-  echo "  (none)"
-else
-  while IFS= read -r line; do
-    [ -z "$line" ] && continue
-    echo "  ✓ $line"
-  done <<< "$SPARE"
-fi
-
-echo ""
-echo "=== In-use worktrees ==="
-if [ -f "$STATE_FILE" ] && jq -e '.' "$STATE_FILE" >/dev/null 2>&1; then
-  IN_USE=$(jq -r '.agents[] | select(.state != "done") | "  [\(.state)] \(.worktree_path) → \(.branch)"' \
-    "$STATE_FILE" 2>/dev/null || echo "")
-  if [ -n "$IN_USE" ]; then
-    echo "$IN_USE"
-  else
-    echo "  (none)"
-  fi
-else
-  echo "  (no active state file)"
-fi

.claude/skills/orchestrate/scripts/classify-pane.sh

@@ -1,85 +0,0 @@
-#!/usr/bin/env bash
-# classify-pane.sh — Classify the current state of a tmux pane
-#
-# Usage: classify-pane.sh <tmux-target>
-#   tmux-target: e.g. "work:0", "work:1.0"
-#
-# Output (stdout): JSON object:
-#   { "state": "running|idle|waiting_approval|complete", "reason": "...", "pane_cmd": "..." }
-#
-# Exit codes: 0=ok, 1=error (invalid target or tmux window not found)
-
-set -euo pipefail
-
-TARGET="${1:-}"
-
-if [ -z "$TARGET" ]; then
-  echo '{"state":"error","reason":"no target provided","pane_cmd":""}'
-  exit 1
-fi
-
-# Validate tmux target format: session:window or session:window.pane
-if ! [[ "$TARGET" =~ ^[a-zA-Z0-9_.-]+:[a-zA-Z0-9_.-]+(\.[0-9]+)?$ ]]; then
-  echo '{"state":"error","reason":"invalid tmux target format","pane_cmd":""}'
-  exit 1
-fi
-
-# Check session exists (use %%:* to extract session name from session:window)
-if ! tmux list-windows -t "${TARGET%%:*}" &>/dev/null 2>&1; then
-  echo '{"state":"error","reason":"tmux target not found","pane_cmd":""}'
-  exit 1
-fi
-
-# Get the current foreground command in the pane
-PANE_CMD=$(tmux display-message -t "$TARGET" -p '#{pane_current_command}' 2>/dev/null || echo "unknown")
-
-# Capture and strip ANSI codes (use perl for cross-platform compatibility — BSD sed lacks \x1b support)
-RAW=$(tmux capture-pane -t "$TARGET" -p -S -50 2>/dev/null || echo "")
-CLEAN=$(echo "$RAW" | perl -pe 's/\x1b\[[0-9;]*[a-zA-Z]//g; s/\x1b\(B//g; s/\x1b\[\?[0-9]*[hl]//g; s/\r//g' \
-  | grep -v '^[[:space:]]*$' || true)
-
-# --- Check: explicit completion marker ---
-# Must be on its own line (not buried in the objective text sent at spawn time).
-if echo "$CLEAN" | grep -qE "^[[:space:]]*ORCHESTRATOR:DONE[[:space:]]*$"; then
-  jq -n --arg cmd "$PANE_CMD" '{"state":"complete","reason":"ORCHESTRATOR:DONE marker found","pane_cmd":$cmd}'
-  exit 0
-fi
-
-# --- Check: Claude Code approval prompt patterns ---
-LAST_40=$(echo "$CLEAN" | tail -40)
-APPROVAL_PATTERNS=(
-  "Do you want to proceed"
-  "Do you want to make this"
-  "\\[y/n\\]"
-  "\\[Y/n\\]"
-  "\\[n/Y\\]"
-  "Proceed\\?"
-  "Allow this command"
-  "Run bash command"
-  "Allow bash"
-  "Would you like"
-  "Press enter to continue"
-  "Esc to cancel"
-)
-for pattern in "${APPROVAL_PATTERNS[@]}"; do
-  if echo "$LAST_40" | grep -qiE "$pattern"; then
-    jq -n --arg pattern "$pattern" --arg cmd "$PANE_CMD" \
-      '{"state":"waiting_approval","reason":"approval pattern: \($pattern)","pane_cmd":$cmd}'
-    exit 0
-  fi
-done
-
-# --- Check: shell prompt (claude has exited) ---
-# If the foreground process is a shell (not claude/node), the agent has exited
-case "$PANE_CMD" in
-  zsh|bash|fish|sh|dash|tcsh|ksh)
-    jq -n --arg cmd "$PANE_CMD" \
-      '{"state":"idle","reason":"agent exited — shell prompt active","pane_cmd":$cmd}'
-    exit 0
-    ;;
-esac
-
-# Agent is still running (claude/node/python is the foreground process)
-jq -n --arg cmd "$PANE_CMD" \
-  '{"state":"running","reason":"foreground process: \($cmd)","pane_cmd":$cmd}'
-exit 0

.claude/skills/orchestrate/scripts/find-spare.sh

@@ -1,24 +0,0 @@
-#!/usr/bin/env bash
-# find-spare.sh — list worktrees on spare/N branches (free to use)
-#
-# Usage: find-spare.sh [REPO_ROOT]
-#   REPO_ROOT defaults to the root worktree containing the current git repo.
-#
-# Output (stdout): one line per available worktree: "PATH BRANCH"
-#   e.g.: /Users/me/Code/AutoGPT3 spare/3
-
-set -euo pipefail
-
-REPO_ROOT="${1:-$(git rev-parse --show-toplevel 2>/dev/null || echo "")}"
-if [ -z "$REPO_ROOT" ]; then
-  echo "Error: not inside a git repo and no REPO_ROOT provided" >&2
-  exit 1
-fi
-
-git -C "$REPO_ROOT" worktree list --porcelain \
-  | awk '
-      /^worktree / { path = substr($0, 10) }
-      /^branch /   { branch = substr($0, 8); print path " " branch }
-    ' \
-  | { grep -E " refs/heads/spare/[0-9]+$" || true; } \
-  | sed 's|refs/heads/||'

.claude/skills/orchestrate/scripts/notify.sh

@@ -1,40 +0,0 @@
-#!/usr/bin/env bash
-# notify.sh — send a fleet notification message
-#
-# Delivery order (first available wins):
-#   1. Discord webhook — DISCORD_WEBHOOK_URL env var OR state file .discord_webhook
-#   2. macOS notification center — osascript (silent fail if unavailable)
-#   3. Stdout only
-#
-# Usage: notify.sh MESSAGE
-# Exit: always 0 (notification failure must not abort the caller)
-
-MESSAGE="${1:-}"
-[ -z "$MESSAGE" ] && exit 0
-
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-
-# --- Resolve Discord webhook ---
-WEBHOOK="${DISCORD_WEBHOOK_URL:-}"
-if [ -z "$WEBHOOK" ] && [ -f "$STATE_FILE" ]; then
-  WEBHOOK=$(jq -r '.discord_webhook // ""' "$STATE_FILE" 2>/dev/null || echo "")
-fi
-
-# --- Discord delivery ---
-if [ -n "$WEBHOOK" ]; then
-  PAYLOAD=$(jq -n --arg msg "$MESSAGE" '{"content": $msg}')
-  curl -s -X POST "$WEBHOOK" \
-    -H "Content-Type: application/json" \
-    -d "$PAYLOAD" > /dev/null 2>&1 || true
-fi
-
-# --- macOS notification center (silent if not macOS or osascript missing) ---
-if command -v osascript &>/dev/null 2>&1; then
-  # Escape single quotes for AppleScript
-  SAFE_MSG=$(echo "$MESSAGE" | sed "s/'/\\\\'/g")
-  osascript -e "display notification \"${SAFE_MSG}\" with title \"Orchestrator\"" 2>/dev/null || true
-fi
-
-# Always print to stdout so run-loop.sh logs it
-echo "$MESSAGE"
-exit 0

.claude/skills/orchestrate/scripts/poll-cycle.sh

@@ -1,257 +0,0 @@
-#!/usr/bin/env bash
-# poll-cycle.sh — Single orchestrator poll cycle
-#
-# Reads ~/.claude/orchestrator-state.json, classifies each agent, updates state,
-# and outputs a JSON array of actions for Claude to take.
-#
-# Usage: poll-cycle.sh
-# Output (stdout): JSON array of action objects
-#   [{ "window": "work:0", "action": "kick|approve|none", "state": "...",
-#      "worktree": "...", "objective": "...", "reason": "..." }]
-#
-# The state file is updated in-place (atomic write via .tmp).
-
-set -euo pipefail
-
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-SCRIPTS_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
-CLASSIFY="$SCRIPTS_DIR/classify-pane.sh"
-
-# Cross-platform md5: always outputs just the hex digest
-md5_hash() {
-  if command -v md5sum &>/dev/null; then
-    md5sum | awk '{print $1}'
-  else
-    md5 | awk '{print $NF}'
-  fi
-}
-
-# Clean up temp file on any exit (avoids stale .tmp if jq write fails)
-trap 'rm -f "${STATE_FILE}.tmp"' EXIT
-
-# Ensure state file exists
-if [ ! -f "$STATE_FILE" ]; then
-  echo '{"active":false,"agents":[]}' > "$STATE_FILE"
-fi
-
-# Validate JSON upfront before any jq reads that run under set -e.
-# A truncated/corrupt file (e.g. from a SIGKILL mid-write) would otherwise
-# abort the script at the ACTIVE read below without emitting any JSON output.
-if ! jq -e '.' "$STATE_FILE" >/dev/null 2>&1; then
-  echo "State file parse error — check $STATE_FILE" >&2
-  echo "[]"
-  exit 0
-fi
-
-ACTIVE=$(jq -r '.active // false' "$STATE_FILE")
-if [ "$ACTIVE" != "true" ]; then
-  echo "[]"
-  exit 0
-fi
-
-NOW=$(date +%s)
-IDLE_THRESHOLD=$(jq -r '.idle_threshold_seconds // 300' "$STATE_FILE")
-
-ACTIONS="[]"
-UPDATED_AGENTS="[]"
-
-# Read agents as newline-delimited JSON objects.
-# jq exits non-zero when .agents[] has no matches on an empty array, which is valid —
-# so we suppress that exit code and separately validate the file is well-formed JSON.
-if ! AGENTS_JSON=$(jq -e -c '.agents // empty | .[]' "$STATE_FILE" 2>/dev/null); then
-  if ! jq -e '.' "$STATE_FILE" > /dev/null 2>&1; then
-    echo "State file parse error — check $STATE_FILE" >&2
-  fi
-  echo "[]"
-  exit 0
-fi
-
-if [ -z "$AGENTS_JSON" ]; then
-  echo "[]"
-  exit 0
-fi
-
-while IFS= read -r agent; do
-  [ -z "$agent" ] && continue
-
-  # Use // "" defaults so a single malformed field doesn't abort the whole cycle
-  WINDOW=$(echo "$agent"   | jq -r '.window // ""')
-  WORKTREE=$(echo "$agent" | jq -r '.worktree // ""')
-  OBJECTIVE=$(echo "$agent"| jq -r '.objective // ""')
-  STATE=$(echo "$agent"    | jq -r '.state // "running"')
-  LAST_HASH=$(echo "$agent"| jq -r '.last_output_hash // ""')
-  IDLE_SINCE=$(echo "$agent"| jq -r '.idle_since // 0')
-  REVISION_COUNT=$(echo "$agent"| jq -r '.revision_count // 0')
-
-  # Validate window format to prevent tmux target injection.
-  # Allow session:window (numeric or named) and session:window.pane
-  if ! [[ "$WINDOW" =~ ^[a-zA-Z0-9_.-]+:[a-zA-Z0-9_.-]+(\.[0-9]+)?$ ]]; then
-    echo "Skipping agent with invalid window value: $WINDOW" >&2
-    UPDATED_AGENTS=$(echo "$UPDATED_AGENTS" | jq --argjson a "$agent" '. + [$a]')
-    continue
-  fi
-
-  # Pass-through terminal-state agents
-  if [[ "$STATE" == "done" || "$STATE" == "escalated" || "$STATE" == "complete" || "$STATE" == "pending_evaluation" ]]; then
-    UPDATED_AGENTS=$(echo "$UPDATED_AGENTS" | jq --argjson a "$agent" '. + [$a]')
-    continue
-  fi
-
-  # Classify pane.
-  # classify-pane.sh always emits JSON before exit (even on error), so using
-  # "|| echo '...'" would concatenate two JSON objects when it exits non-zero.
-  # Use "|| true" inside the substitution so set -euo pipefail does not abort
-  # the poll cycle when classify exits with a non-zero status code.
-  CLASSIFICATION=$("$CLASSIFY" "$WINDOW" 2>/dev/null || true)
-  [ -z "$CLASSIFICATION" ] && CLASSIFICATION='{"state":"error","reason":"classify failed","pane_cmd":"unknown"}'
-
-  PANE_STATE=$(echo "$CLASSIFICATION" | jq -r '.state')
-  PANE_REASON=$(echo "$CLASSIFICATION" | jq -r '.reason')
-
-  # Capture full pane output once — used for hash (stuck detection) and checkpoint parsing.
-  # Use -S -500 to get the last ~500 lines of scrollback so checkpoints aren't missed.
-  RAW=$(tmux capture-pane -t "$WINDOW" -p -S -500 2>/dev/null || echo "")
-
-  # --- Checkpoint tracking ---
-  # Parse any "CHECKPOINT:<step>" lines the agent has output and merge into state file.
-  # The agent writes these as it completes each required step so verify-complete.sh can gate recycling.
-  EXISTING_CPS=$(echo "$agent" | jq -c '.checkpoints // []')
-  NEW_CHECKPOINTS_JSON="$EXISTING_CPS"
-  if [ -n "$RAW" ]; then
-    FOUND_CPS=$(echo "$RAW" \
-      | grep -oE "CHECKPOINT:[a-zA-Z0-9_-]+" \
-      | sed 's/CHECKPOINT://' \
-      | sort -u \
-      | jq -R . | jq -s . 2>/dev/null || echo "[]")
-    NEW_CHECKPOINTS_JSON=$(jq -n \
-      --argjson existing "$EXISTING_CPS" \
-      --argjson found "$FOUND_CPS" \
-      '($existing + $found) | unique' 2>/dev/null || echo "$EXISTING_CPS")
-  fi
-
-  # Compute content hash for stuck-detection (only for running agents)
-  CURRENT_HASH=""
-  if [[ "$PANE_STATE" == "running" ]] && [ -n "$RAW" ]; then
-    CURRENT_HASH=$(echo "$RAW" | tail -20 | md5_hash)
-  fi
-
-  NEW_STATE="$STATE"
-  NEW_IDLE_SINCE="$IDLE_SINCE"
-  NEW_REVISION_COUNT="$REVISION_COUNT"
-  ACTION="none"
-  REASON="$PANE_REASON"
-
-  case "$PANE_STATE" in
-    complete)
-      # Agent output ORCHESTRATOR:DONE — mark pending_evaluation so orchestrator handles it.
-      # run-loop does NOT verify or notify; orchestrator's background poll picks this up.
-      NEW_STATE="pending_evaluation"
-      ACTION="complete"  # run-loop logs it but takes no action
-      ;;
-    waiting_approval)
-      NEW_STATE="waiting_approval"
-      ACTION="approve"
-      ;;
-    idle)
-      # Agent process has exited — needs restart
-      NEW_STATE="idle"
-      ACTION="kick"
-      REASON="agent exited (shell is foreground)"
-      NEW_REVISION_COUNT=$(( REVISION_COUNT + 1 ))
-      NEW_IDLE_SINCE=$NOW
-      if [ "$NEW_REVISION_COUNT" -ge 3 ]; then
-        NEW_STATE="escalated"
-        ACTION="none"
-        REASON="escalated after ${NEW_REVISION_COUNT} kicks — needs human attention"
-      fi
-      ;;
-    running)
-      # Clear idle_since only when transitioning from idle (agent was kicked and
-      # restarted). Do NOT reset for stuck — idle_since must persist across polls
-      # so STUCK_DURATION can accumulate and trigger escalation.
-      # Also update the local IDLE_SINCE so the hash-stability check below uses
-      # the reset value on this same poll, not the stale kick timestamp.
-      if [[ "$STATE" == "idle" ]]; then
-        NEW_IDLE_SINCE=0
-        IDLE_SINCE=0
-      fi
-      # Check if hash has been stable (agent may be stuck mid-task)
-      if [ -n "$CURRENT_HASH" ] && [ "$CURRENT_HASH" = "$LAST_HASH" ] && [ "$LAST_HASH" != "" ]; then
-        if [ "$IDLE_SINCE" = "0" ] || [ "$IDLE_SINCE" = "null" ]; then
-          NEW_IDLE_SINCE=$NOW
-        else
-          STUCK_DURATION=$(( NOW - IDLE_SINCE ))
-          if [ "$STUCK_DURATION" -gt "$IDLE_THRESHOLD" ]; then
-            NEW_REVISION_COUNT=$(( REVISION_COUNT + 1 ))
-            NEW_IDLE_SINCE=$NOW
-            if [ "$NEW_REVISION_COUNT" -ge 3 ]; then
-              NEW_STATE="escalated"
-              ACTION="none"
-              REASON="escalated after ${NEW_REVISION_COUNT} kicks — needs human attention"
-            else
-              NEW_STATE="stuck"
-              ACTION="kick"
-              REASON="output unchanged for ${STUCK_DURATION}s (threshold: ${IDLE_THRESHOLD}s)"
-            fi
-          fi
-        fi
-      else
-        # Only reset the idle timer when we have a valid hash comparison (pane
-        # capture succeeded). If CURRENT_HASH is empty (tmux capture-pane failed),
-        # preserve existing timers so stuck detection is not inadvertently reset.
-        if [ -n "$CURRENT_HASH" ]; then
-          NEW_STATE="running"
-          NEW_IDLE_SINCE=0
-        fi
-      fi
-      ;;
-    error)
-      REASON="classify error: $PANE_REASON"
-      ;;
-  esac
-
-  # Build updated agent record (ensure idle_since and revision_count are numeric)
-  # Use || true on each jq call so a malformed field skips this agent rather than
-  # aborting the entire poll cycle under set -e.
-  UPDATED_AGENT=$(echo "$agent" | jq \
-    --arg state "$NEW_STATE" \
-    --arg hash "$CURRENT_HASH" \
-    --argjson now "$NOW" \
-    --arg idle_since "$NEW_IDLE_SINCE" \
-    --arg revision_count "$NEW_REVISION_COUNT" \
-    --argjson checkpoints "$NEW_CHECKPOINTS_JSON" \
-    '.state = $state
-     | .last_output_hash = (if $hash == "" then .last_output_hash else $hash end)
-     | .last_seen_at = $now
-     | .idle_since = ($idle_since | tonumber)
-     | .revision_count = ($revision_count | tonumber)
-     | .checkpoints = $checkpoints' 2>/dev/null) || {
-    echo "Warning: failed to build updated agent for window $WINDOW — keeping original" >&2
-    UPDATED_AGENTS=$(echo "$UPDATED_AGENTS" | jq --argjson a "$agent" '. + [$a]')
-    continue
-  }
-
-  UPDATED_AGENTS=$(echo "$UPDATED_AGENTS" | jq --argjson a "$UPDATED_AGENT" '. + [$a]')
-
-  # Add action if needed
-  if [ "$ACTION" != "none" ]; then
-    ACTION_OBJ=$(jq -n \
-      --arg window "$WINDOW" \
-      --arg action "$ACTION" \
-      --arg state "$NEW_STATE" \
-      --arg worktree "$WORKTREE" \
-      --arg objective "$OBJECTIVE" \
-      --arg reason "$REASON" \
-      '{window:$window, action:$action, state:$state, worktree:$worktree, objective:$objective, reason:$reason}')
-    ACTIONS=$(echo "$ACTIONS" | jq --argjson a "$ACTION_OBJ" '. + [$a]')
-  fi
-
-done <<< "$AGENTS_JSON"
-
-# Atomic state file update
-jq --argjson agents "$UPDATED_AGENTS" \
-   --argjson now "$NOW" \
-   '.agents = $agents | .last_poll_at = $now' \
-   "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
-
-echo "$ACTIONS"

.claude/skills/orchestrate/scripts/recycle-agent.sh

@@ -1,32 +0,0 @@
-#!/usr/bin/env bash
-# recycle-agent.sh — kill a tmux window and restore the worktree to its spare branch
-#
-# Usage: recycle-agent.sh WINDOW WORKTREE_PATH SPARE_BRANCH
-#   WINDOW        — tmux target, e.g. autogpt1:3
-#   WORKTREE_PATH — absolute path to the git worktree
-#   SPARE_BRANCH  — branch to restore, e.g. spare/6
-#
-# Stdout: one status line
-
-set -euo pipefail
-
-if [ $# -lt 3 ]; then
-  echo "Usage: recycle-agent.sh WINDOW WORKTREE_PATH SPARE_BRANCH" >&2
-  exit 1
-fi
-
-WINDOW="$1"
-WORKTREE_PATH="$2"
-SPARE_BRANCH="$3"
-
-# Kill the tmux window (ignore error — may already be gone)
-tmux kill-window -t "$WINDOW" 2>/dev/null || true
-
-# Restore to spare branch: abort any in-progress operation, then clean
-git -C "$WORKTREE_PATH" rebase --abort 2>/dev/null || true
-git -C "$WORKTREE_PATH" merge --abort 2>/dev/null || true
-git -C "$WORKTREE_PATH" reset --hard HEAD 2>/dev/null
-git -C "$WORKTREE_PATH" clean -fd 2>/dev/null
-git -C "$WORKTREE_PATH" checkout "$SPARE_BRANCH"
-
-echo "Recycled: $(basename "$WORKTREE_PATH") → $SPARE_BRANCH (window $WINDOW closed)"

.claude/skills/orchestrate/scripts/run-loop.sh

@@ -1,164 +0,0 @@
-#!/usr/bin/env bash
-# run-loop.sh — Mechanical babysitter for the agent fleet (runs in its own tmux window)
-#
-# Handles ONLY two things that need no intelligence:
-#   idle    → restart claude using --resume SESSION_ID (or --continue) to restore context
-#   approve → auto-approve safe dialogs, press Enter on numbered-option dialogs
-#
-# Everything else — ORCHESTRATOR:DONE, verification, /pr-test, final evaluation,
-# marking done, deciding to close windows — is the orchestrating Claude's job.
-# poll-cycle.sh sets state to pending_evaluation when ORCHESTRATOR:DONE is detected;
-# the orchestrator's background poll loop handles it from there.
-#
-# Usage: run-loop.sh
-# Env:   POLL_INTERVAL (default: 30), ORCHESTRATOR_STATE_FILE
-
-set -euo pipefail
-
-# Copy scripts to a stable location outside the repo so they survive branch
-# checkouts (e.g. recycle-agent.sh switching spare/N back into this worktree
-# would wipe .claude/skills/orchestrate/scripts if the skill only exists on the
-# current branch).
-_ORIGIN_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
-STABLE_SCRIPTS_DIR="$HOME/.claude/orchestrator/scripts"
-mkdir -p "$STABLE_SCRIPTS_DIR"
-cp "$_ORIGIN_DIR"/*.sh "$STABLE_SCRIPTS_DIR/"
-chmod +x "$STABLE_SCRIPTS_DIR"/*.sh
-SCRIPTS_DIR="$STABLE_SCRIPTS_DIR"
-
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-POLL_INTERVAL="${POLL_INTERVAL:-30}"
-
-# ---------------------------------------------------------------------------
-# update_state WINDOW FIELD VALUE
-# ---------------------------------------------------------------------------
-update_state() {
-  local window="$1" field="$2" value="$3"
-  jq --arg w "$window" --arg f "$field" --arg v "$value" \
-    '.agents |= map(if .window == $w then .[$f] = $v else . end)' \
-    "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
-}
-
-update_state_int() {
-  local window="$1" field="$2" value="$3"
-  jq --arg w "$window" --arg f "$field" --argjson v "$value" \
-    '.agents |= map(if .window == $w then .[$f] = $v else . end)' \
-    "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
-}
-
-agent_field() {
-  jq -r --arg w "$1" --arg f "$2" \
-    '.agents[] | select(.window == $w) | .[$f] // ""' \
-    "$STATE_FILE" 2>/dev/null
-}
-
-# ---------------------------------------------------------------------------
-# wait_for_prompt WINDOW — wait up to 60s for Claude's ❯ prompt
-# ---------------------------------------------------------------------------
-wait_for_prompt() {
-  local window="$1"
-  for i in $(seq 1 60); do
-    local cmd pane
-    cmd=$(tmux display-message -t "$window" -p '#{pane_current_command}' 2>/dev/null || echo "")
-    pane=$(tmux capture-pane -t "$window" -p 2>/dev/null || echo "")
-    if echo "$pane" | grep -q "Enter to confirm"; then
-      tmux send-keys -t "$window" Down Enter; sleep 2; continue
-    fi
-    [[ "$cmd" == "node" ]] && echo "$pane" | grep -q "❯" && return 0
-    sleep 1
-  done
-  return 1  # timed out
-}
-
-# ---------------------------------------------------------------------------
-# handle_kick WINDOW STATE — only for idle (crashed) agents, not stuck
-# ---------------------------------------------------------------------------
-handle_kick() {
-  local window="$1" state="$2"
-  [[ "$state" != "idle" ]] && return  # stuck agents handled by supervisor
-
-  local worktree_path session_id
-  worktree_path=$(agent_field "$window" "worktree_path")
-  session_id=$(agent_field "$window" "session_id")
-
-  echo "[$(date +%H:%M:%S)] KICK restart  $window — agent exited, resuming session"
-
-  # Resume the exact session so the agent retains full context — no need to re-send objective
-  if [ -n "$session_id" ]; then
-    tmux send-keys -t "$window" "cd '${worktree_path}' && claude --resume '${session_id}' --permission-mode bypassPermissions" Enter
-  else
-    tmux send-keys -t "$window" "cd '${worktree_path}' && claude --continue --permission-mode bypassPermissions" Enter
-  fi
-
-  wait_for_prompt "$window" || echo "[$(date +%H:%M:%S)] KICK WARNING  $window — timed out waiting for ❯"
-}
-
-# ---------------------------------------------------------------------------
-# handle_approve WINDOW — auto-approve dialogs that need no judgment
-# ---------------------------------------------------------------------------
-handle_approve() {
-  local window="$1"
-  local pane_tail
-  pane_tail=$(tmux capture-pane -t "$window" -p 2>/dev/null | tail -3 || echo "")
-
-  # Settings error dialog at startup
-  if echo "$pane_tail" | grep -q "Enter to confirm"; then
-    echo "[$(date +%H:%M:%S)] APPROVE dialog $window — settings error"
-    tmux send-keys -t "$window" Down Enter
-    return
-  fi
-
-  # Numbered-option dialog (e.g. "Do you want to make this edit?")
-  # ❯ is already on option 1 (Yes) — Enter confirms it
-  if echo "$pane_tail" | grep -qE "❯\s*1\." || echo "$pane_tail" | grep -q "Esc to cancel"; then
-    echo "[$(date +%H:%M:%S)] APPROVE edit   $window"
-    tmux send-keys -t "$window" "" Enter
-    return
-  fi
-
-  # y/n prompt for safe operations
-  if echo "$pane_tail" | grep -qiE "(^git |^npm |^pnpm |^poetry |^pytest|^docker |^make |^cargo |^pip |^yarn |curl .*(localhost|127\.0\.0\.1))"; then
-    echo "[$(date +%H:%M:%S)] APPROVE safe   $window"
-    tmux send-keys -t "$window" "y" Enter
-    return
-  fi
-
-  # Anything else — supervisor handles it, just log
-  echo "[$(date +%H:%M:%S)] APPROVE skip   $window — unknown dialog, supervisor will handle"
-}
-
-# ---------------------------------------------------------------------------
-# Main loop
-# ---------------------------------------------------------------------------
-echo "[$(date +%H:%M:%S)] run-loop started (mechanical only, poll every ${POLL_INTERVAL}s)"
-echo "[$(date +%H:%M:%S)] Supervisor: orchestrating Claude session (not a separate window)"
-echo "---"
-
-while true; do
-  if ! jq -e '.active == true' "$STATE_FILE" >/dev/null 2>&1; then
-    echo "[$(date +%H:%M:%S)] active=false — exiting."
-    exit 0
-  fi
-
-  ACTIONS=$("$SCRIPTS_DIR/poll-cycle.sh" 2>/dev/null || echo "[]")
-  KICKED=0; DONE=0
-
-  while IFS= read -r action; do
-    [ -z "$action" ] && continue
-    WINDOW=$(echo "$action" | jq -r '.window // ""')
-    ACTION=$(echo "$action" | jq -r '.action // ""')
-    STATE=$(echo "$action"  | jq -r '.state // ""')
-
-    case "$ACTION" in
-      kick)     handle_kick "$WINDOW" "$STATE" || true; KICKED=$(( KICKED + 1 )) ;;
-      approve)  handle_approve "$WINDOW" || true ;;
-      complete) DONE=$(( DONE + 1 )) ;;  # poll-cycle already set state=pending_evaluation; orchestrator handles
-    esac
-  done < <(echo "$ACTIONS" | jq -c '.[]' 2>/dev/null || true)
-
-  RUNNING=$(jq '[.agents[] | select(.state | test("running|stuck|waiting_approval|idle"))] | length' \
-    "$STATE_FILE" 2>/dev/null || echo 0)
-
-  echo "[$(date +%H:%M:%S)] Poll — ${RUNNING} running  ${KICKED} kicked  ${DONE} recycled"
-  sleep "$POLL_INTERVAL"
-done

.claude/skills/orchestrate/scripts/spawn-agent.sh

@@ -1,122 +0,0 @@
-#!/usr/bin/env bash
-# spawn-agent.sh — create tmux window, checkout branch, launch claude, send task
-#
-# Usage: spawn-agent.sh SESSION WORKTREE_PATH SPARE_BRANCH NEW_BRANCH OBJECTIVE [PR_NUMBER] [STEPS...]
-#   SESSION       — tmux session name, e.g. autogpt1
-#   WORKTREE_PATH — absolute path to the git worktree
-#   SPARE_BRANCH  — spare branch being replaced, e.g. spare/6 (saved for recycle)
-#   NEW_BRANCH    — task branch to create, e.g. feat/my-feature
-#   OBJECTIVE     — task description sent to the agent
-#   PR_NUMBER     — (optional) GitHub PR number for completion verification
-#   STEPS...      — (optional) required checkpoint names, e.g. pr-address pr-test
-#
-# Stdout: SESSION:WINDOW_INDEX (nothing else — callers rely on this)
-# Exit non-zero on failure.
-
-set -euo pipefail
-
-if [ $# -lt 5 ]; then
-  echo "Usage: spawn-agent.sh SESSION WORKTREE_PATH SPARE_BRANCH NEW_BRANCH OBJECTIVE [PR_NUMBER] [STEPS...]" >&2
-  exit 1
-fi
-
-SESSION="$1"
-WORKTREE_PATH="$2"
-SPARE_BRANCH="$3"
-NEW_BRANCH="$4"
-OBJECTIVE="$5"
-PR_NUMBER="${6:-}"
-STEPS=("${@:7}")
-WORKTREE_NAME=$(basename "$WORKTREE_PATH")
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-
-# Generate a stable session ID so this agent's Claude session can always be resumed:
-#   claude --resume $SESSION_ID --permission-mode bypassPermissions
-SESSION_ID=$(uuidgen 2>/dev/null || python3 -c "import uuid; print(uuid.uuid4())")
-
-# Create (or switch to) the task branch
-git -C "$WORKTREE_PATH" checkout -b "$NEW_BRANCH" 2>/dev/null \
-  || git -C "$WORKTREE_PATH" checkout "$NEW_BRANCH"
-
-# Open a new named tmux window; capture its numeric index
-WIN_IDX=$(tmux new-window -t "$SESSION" -n "$WORKTREE_NAME" -P -F '#{window_index}')
-WINDOW="${SESSION}:${WIN_IDX}"
-
-# Append the initial agent record to the state file so subsequent jq updates find it.
-# This must happen before the pr_number/steps update below.
-if [ -f "$STATE_FILE" ]; then
-  NOW=$(date +%s)
-  jq --arg window "$WINDOW" \
-     --arg worktree "$WORKTREE_NAME" \
-     --arg worktree_path "$WORKTREE_PATH" \
-     --arg spare_branch "$SPARE_BRANCH" \
-     --arg branch "$NEW_BRANCH" \
-     --arg objective "$OBJECTIVE" \
-     --arg session_id "$SESSION_ID" \
-     --argjson now "$NOW" \
-     '.agents += [{
-       "window": $window,
-       "worktree": $worktree,
-       "worktree_path": $worktree_path,
-       "spare_branch": $spare_branch,
-       "branch": $branch,
-       "objective": $objective,
-       "session_id": $session_id,
-       "state": "running",
-       "checkpoints": [],
-       "last_output_hash": "",
-       "last_seen_at": $now,
-       "spawned_at": $now,
-       "idle_since": 0,
-       "revision_count": 0,
-       "last_rebriefed_at": 0
-     }]' "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
-fi
-
-# Store pr_number + steps in state file if provided (enables verify-complete.sh).
-# The agent record was appended above so the jq select now finds it.
-if [ -n "$PR_NUMBER" ] && [ -f "$STATE_FILE" ]; then
-  if [ "${#STEPS[@]}" -gt 0 ]; then
-    STEPS_JSON=$(printf '%s\n' "${STEPS[@]}" | jq -R . | jq -s .)
-  else
-    STEPS_JSON='[]'
-  fi
-  jq --arg w "$WINDOW" --arg pr "$PR_NUMBER" --argjson steps "$STEPS_JSON" \
-    '.agents |= map(if .window == $w then . + {pr_number: $pr, steps: $steps, checkpoints: []} else . end)' \
-    "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
-fi
-
-# Launch claude with a stable session ID so it can always be resumed after a crash:
-#   claude --resume SESSION_ID --permission-mode bypassPermissions
-tmux send-keys -t "$WINDOW" "cd '${WORKTREE_PATH}' && claude --permission-mode bypassPermissions --session-id '${SESSION_ID}'" Enter
-
-# Wait up to 60s for claude to be fully interactive:
-# both pane_current_command == 'node' AND the '❯' prompt is visible.
-PROMPT_FOUND=false
-for i in $(seq 1 60); do
-  CMD=$(tmux display-message -t "$WINDOW" -p '#{pane_current_command}' 2>/dev/null || echo "")
-  PANE=$(tmux capture-pane -t "$WINDOW" -p 2>/dev/null || echo "")
-  if echo "$PANE" | grep -q "Enter to confirm"; then
-    tmux send-keys -t "$WINDOW" Down Enter
-    sleep 2
-    continue
-  fi
-  if [[ "$CMD" == "node" ]] && echo "$PANE" | grep -q "❯"; then
-    PROMPT_FOUND=true
-    break
-  fi
-  sleep 1
-done
-
-if ! $PROMPT_FOUND; then
-  echo "[spawn-agent] WARNING: timed out waiting for ❯ prompt on $WINDOW — sending objective anyway" >&2
-fi
-
-# Send the task. Split text and Enter — if combined, Enter can fire before the string
-# is fully buffered, leaving the message stuck as "[Pasted text +N lines]" unsent.
-tmux send-keys -t "$WINDOW" "${OBJECTIVE} Output each completed step as CHECKPOINT:<step-name>. When ALL steps are done, output ORCHESTRATOR:DONE on its own line."
-sleep 0.3
-tmux send-keys -t "$WINDOW" Enter
-
-# Only output the window address — nothing else (callers parse this)
-echo "$WINDOW"

.claude/skills/orchestrate/scripts/status.sh

@@ -1,43 +0,0 @@
-#!/usr/bin/env bash
-# status.sh — print orchestrator status: state file summary + live tmux pane commands
-#
-# Usage: status.sh
-# Reads: ~/.claude/orchestrator-state.json
-
-set -euo pipefail
-
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-
-if [ ! -f "$STATE_FILE" ] || ! jq -e '.' "$STATE_FILE" >/dev/null 2>&1; then
-  echo "No orchestrator state found at $STATE_FILE"
-  exit 0
-fi
-
-# Header: active status, session, thresholds, last poll
-jq -r '
-  "=== Orchestrator [\(if .active then "RUNNING" else "STOPPED" end)] ===",
-  "Session: \(.tmux_session // "unknown")  |  Idle threshold: \(.idle_threshold_seconds // 300)s",
-  "Last poll: \(if (.last_poll_at // 0) == 0 then "never" else (.last_poll_at | strftime("%H:%M:%S")) end)",
-  ""
-' "$STATE_FILE"
-
-# Each agent: state, window, worktree/branch, truncated objective
-AGENT_COUNT=$(jq '.agents | length' "$STATE_FILE")
-if [ "$AGENT_COUNT" -eq 0 ]; then
-  echo "  (no agents registered)"
-else
-  jq -r '
-    .agents[] |
-    "  [\(.state | ascii_upcase)] \(.window)  \(.worktree)/\(.branch)",
-    "    \(.objective // "" | .[0:70])"
-  ' "$STATE_FILE"
-fi
-
-echo ""
-
-# Live pane_current_command for non-done agents
-while IFS= read -r WINDOW; do
-  [ -z "$WINDOW" ] && continue
-  CMD=$(tmux display-message -t "$WINDOW" -p '#{pane_current_command}' 2>/dev/null || echo "unreachable")
-  echo "  $WINDOW live: $CMD"
-done < <(jq -r '.agents[] | select(.state != "done") | .window' "$STATE_FILE" 2>/dev/null || true)

.claude/skills/orchestrate/scripts/verify-complete.sh

@@ -1,180 +0,0 @@
-#!/usr/bin/env bash
-# verify-complete.sh — verify a PR task is truly done before marking the agent done
-#
-# Check order matters:
-#   1. Checkpoints — did the agent do all required steps?
-#   2. CI complete — no pending (bots post comments AFTER their check runs, must wait)
-#   3. CI passing — no failures (agent must fix before done)
-#   4. spawned_at — a new CI run was triggered after agent spawned (proves real work)
-#   5. Unresolved threads — checked AFTER CI so bot-posted comments are included
-#   6. CHANGES_REQUESTED — checked AFTER CI so bot reviews are included
-#
-# Usage: verify-complete.sh WINDOW
-# Exit 0 = verified complete; exit 1 = not complete (stderr has reason)
-
-set -euo pipefail
-
-WINDOW="$1"
-STATE_FILE="${ORCHESTRATOR_STATE_FILE:-$HOME/.claude/orchestrator-state.json}"
-
-PR_NUMBER=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .pr_number // ""' "$STATE_FILE" 2>/dev/null)
-STEPS=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .steps // [] | .[]' "$STATE_FILE" 2>/dev/null || true)
-CHECKPOINTS=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .checkpoints // [] | .[]' "$STATE_FILE" 2>/dev/null || true)
-WORKTREE_PATH=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .worktree_path // ""' "$STATE_FILE" 2>/dev/null)
-BRANCH=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .branch // ""' "$STATE_FILE" 2>/dev/null)
-SPAWNED_AT=$(jq -r --arg w "$WINDOW" '.agents[] | select(.window == $w) | .spawned_at // "0"' "$STATE_FILE" 2>/dev/null || echo "0")
-
-# No PR number = cannot verify
-if [ -z "$PR_NUMBER" ]; then
-  echo "NOT COMPLETE: no pr_number in state — set pr_number or mark done manually" >&2
-  exit 1
-fi
-
-# --- Check 1: all required steps are checkpointed ---
-MISSING=""
-while IFS= read -r step; do
-  [ -z "$step" ] && continue
-  if ! echo "$CHECKPOINTS" | grep -qFx "$step"; then
-    MISSING="$MISSING $step"
-  fi
-done <<< "$STEPS"
-
-if [ -n "$MISSING" ]; then
-  echo "NOT COMPLETE: missing checkpoints:$MISSING on PR #$PR_NUMBER" >&2
-  exit 1
-fi
-
-# Resolve repo for all GitHub checks below
-REPO=$(jq -r '.repo // ""' "$STATE_FILE" 2>/dev/null || echo "")
-if [ -z "$REPO" ] && [ -n "$WORKTREE_PATH" ] && [ -d "$WORKTREE_PATH" ]; then
-  REPO=$(git -C "$WORKTREE_PATH" remote get-url origin 2>/dev/null \
-    | sed 's|.*github\.com[:/]||; s|\.git$||' || echo "")
-fi
-
-if [ -z "$REPO" ]; then
-  echo "Warning: cannot resolve repo — skipping CI/thread checks" >&2
-  echo "VERIFIED: PR #$PR_NUMBER — checkpoints ✓ (CI/thread checks skipped — no repo)"
-  exit 0
-fi
-
-CI_BUCKETS=$(gh pr checks "$PR_NUMBER" --repo "$REPO" --json bucket 2>/dev/null || echo "[]")
-
-# --- Check 2: CI fully complete — no pending checks ---
-# Pending checks MUST finish before we check threads/reviews:
-# bots (Seer, Check PR Status, etc.) post comments and CHANGES_REQUESTED AFTER their CI check runs.
-PENDING=$(echo "$CI_BUCKETS" | jq '[.[] | select(.bucket == "pending")] | length' 2>/dev/null || echo "0")
-if [ "$PENDING" -gt 0 ]; then
-  PENDING_NAMES=$(gh pr checks "$PR_NUMBER" --repo "$REPO" --json bucket,name 2>/dev/null \
-    | jq -r '[.[] | select(.bucket == "pending") | .name] | join(", ")' 2>/dev/null || echo "unknown")
-  echo "NOT COMPLETE: $PENDING CI checks still pending on PR #$PR_NUMBER ($PENDING_NAMES)" >&2
-  exit 1
-fi
-
-# --- Check 3: CI passing — no failures ---
-FAILING=$(echo "$CI_BUCKETS" | jq '[.[] | select(.bucket == "fail")] | length' 2>/dev/null || echo "0")
-if [ "$FAILING" -gt 0 ]; then
-  FAILING_NAMES=$(gh pr checks "$PR_NUMBER" --repo "$REPO" --json bucket,name 2>/dev/null \
-    | jq -r '[.[] | select(.bucket == "fail") | .name] | join(", ")' 2>/dev/null || echo "unknown")
-  echo "NOT COMPLETE: $FAILING failing CI checks on PR #$PR_NUMBER ($FAILING_NAMES)" >&2
-  exit 1
-fi
-
-# --- Check 4: a new CI run was triggered AFTER the agent spawned ---
-if [ -n "$BRANCH" ] && [ "${SPAWNED_AT:-0}" -gt 0 ]; then
-  LATEST_RUN_AT=$(gh run list --repo "$REPO" --branch "$BRANCH" \
-    --json createdAt --limit 1 2>/dev/null | jq -r '.[0].createdAt // ""')
-  if [ -n "$LATEST_RUN_AT" ]; then
-    if date --version >/dev/null 2>&1; then
-      LATEST_RUN_EPOCH=$(date -d "$LATEST_RUN_AT" "+%s" 2>/dev/null || echo "0")
-    else
-      LATEST_RUN_EPOCH=$(TZ=UTC date -j -f "%Y-%m-%dT%H:%M:%SZ" "$LATEST_RUN_AT" "+%s" 2>/dev/null || echo "0")
-    fi
-    if [ "$LATEST_RUN_EPOCH" -le "$SPAWNED_AT" ]; then
-      echo "NOT COMPLETE: latest CI run on $BRANCH predates agent spawn — agent may not have pushed yet" >&2
-      exit 1
-    fi
-  fi
-fi
-
-OWNER=$(echo "$REPO" | cut -d/ -f1)
-REPONAME=$(echo "$REPO" | cut -d/ -f2)
-
-# --- Check 5: no unresolved review threads (checked AFTER CI — bots post after their check) ---
-UNRESOLVED=$(gh api graphql -f query="
-  { repository(owner: \"${OWNER}\", name: \"${REPONAME}\") {
-      pullRequest(number: ${PR_NUMBER}) {
-        reviewThreads(first: 50) { nodes { isResolved } }
-      }
-    }
-  }
-" --jq '[.data.repository.pullRequest.reviewThreads.nodes[] | select(.isResolved == false)] | length' 2>/dev/null || echo "0")
-
-if [ "$UNRESOLVED" -gt 0 ]; then
-  echo "NOT COMPLETE: $UNRESOLVED unresolved review threads on PR #$PR_NUMBER" >&2
-  exit 1
-fi
-
-# --- Check 6: no CHANGES_REQUESTED (checked AFTER CI — bots post reviews after their check) ---
-# A CHANGES_REQUESTED review is stale if the latest commit was pushed AFTER the review was submitted.
-# Stale reviews (pre-dating the fixing commits) should not block verification.
-#
-# Fetch commits and latestReviews in a single call and fail closed — if gh fails,
-# treat that as NOT COMPLETE rather than silently passing.
-# Use latestReviews (not reviews) so each reviewer's latest state is used — superseded
-# CHANGES_REQUESTED entries are automatically excluded when the reviewer later approved.
-# Note: we intentionally use committedDate (not PR updatedAt) because updatedAt changes on any
-# PR activity (bot comments, label changes) which would create false negatives.
-PR_REVIEW_METADATA=$(gh pr view "$PR_NUMBER" --repo "$REPO" \
-  --json commits,latestReviews 2>/dev/null) || {
-  echo "NOT COMPLETE: unable to fetch PR review metadata for PR #$PR_NUMBER" >&2
-  exit 1
-}
-
-LATEST_COMMIT_DATE=$(jq -r '.commits[-1].committedDate // ""' <<< "$PR_REVIEW_METADATA")
-CHANGES_REQUESTED_REVIEWS=$(jq '[.latestReviews[]? | select(.state == "CHANGES_REQUESTED")]' <<< "$PR_REVIEW_METADATA")
-
-BLOCKING_CHANGES_REQUESTED=0
-BLOCKING_REQUESTERS=""
-
-if [ -n "$LATEST_COMMIT_DATE" ] && [ "$(echo "$CHANGES_REQUESTED_REVIEWS" | jq length)" -gt 0 ]; then
-  if date --version >/dev/null 2>&1; then
-    LATEST_COMMIT_EPOCH=$(date -d "$LATEST_COMMIT_DATE" "+%s" 2>/dev/null || echo "0")
-  else
-    LATEST_COMMIT_EPOCH=$(TZ=UTC date -j -f "%Y-%m-%dT%H:%M:%SZ" "$LATEST_COMMIT_DATE" "+%s" 2>/dev/null || echo "0")
-  fi
-
-  while IFS= read -r review; do
-    [ -z "$review" ] && continue
-    REVIEW_DATE=$(echo "$review" | jq -r '.submittedAt // ""')
-    REVIEWER=$(echo "$review" | jq -r '.author.login // "unknown"')
-    if [ -z "$REVIEW_DATE" ]; then
-      # No submission date — treat as fresh (conservative: blocks verification)
-      BLOCKING_CHANGES_REQUESTED=$(( BLOCKING_CHANGES_REQUESTED + 1 ))
-      BLOCKING_REQUESTERS="${BLOCKING_REQUESTERS:+$BLOCKING_REQUESTERS, }${REVIEWER}"
-    else
-      if date --version >/dev/null 2>&1; then
-        REVIEW_EPOCH=$(date -d "$REVIEW_DATE" "+%s" 2>/dev/null || echo "0")
-      else
-        REVIEW_EPOCH=$(TZ=UTC date -j -f "%Y-%m-%dT%H:%M:%SZ" "$REVIEW_DATE" "+%s" 2>/dev/null || echo "0")
-      fi
-      if [ "$REVIEW_EPOCH" -gt "$LATEST_COMMIT_EPOCH" ]; then
-        # Review was submitted AFTER latest commit — still fresh, blocks verification
-        BLOCKING_CHANGES_REQUESTED=$(( BLOCKING_CHANGES_REQUESTED + 1 ))
-        BLOCKING_REQUESTERS="${BLOCKING_REQUESTERS:+$BLOCKING_REQUESTERS, }${REVIEWER}"
-      fi
-      # Review submitted BEFORE latest commit — stale, skip
-    fi
-  done <<< "$(echo "$CHANGES_REQUESTED_REVIEWS" | jq -c '.[]')"
-else
-  # No commit date or no changes_requested — check raw count as fallback
-  BLOCKING_CHANGES_REQUESTED=$(echo "$CHANGES_REQUESTED_REVIEWS" | jq length 2>/dev/null || echo "0")
-  BLOCKING_REQUESTERS=$(echo "$CHANGES_REQUESTED_REVIEWS" | jq -r '[.[].author.login] | join(", ")' 2>/dev/null || echo "unknown")
-fi
-
-if [ "$BLOCKING_CHANGES_REQUESTED" -gt 0 ]; then
-  echo "NOT COMPLETE: CHANGES_REQUESTED (after latest commit) from ${BLOCKING_REQUESTERS} on PR #$PR_NUMBER" >&2
-  exit 1
-fi
-
-echo "VERIFIED: PR #$PR_NUMBER — checkpoints ✓, CI complete + green, 0 unresolved threads, no CHANGES_REQUESTED"
-exit 0

.claude/skills/pr-address/SKILL.md

@@ -90,34 +90,10 @@ Address comments **one at a time**: fix → commit → push → inline reply →
 2. Commit and push the fix
 3. Reply **inline** (not as a new top-level comment) referencing the fixing commit — this is what resolves the conversation for bot reviewers (coderabbitai, sentry):
 
-Use a **markdown commit link** so GitHub renders it as a clickable reference. Get the full SHA with `git rev-parse HEAD` after committing:
-
 | Comment type | How to reply |
 |---|---|
-| Inline review (`pulls/{N}/comments`) | `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/comments/{ID}/replies -f body="🤖 Fixed in [abc1234](https://github.com/Significant-Gravitas/AutoGPT/commit/FULL_SHA): <description>"` |
-| Conversation (`issues/{N}/comments`) | `gh api repos/Significant-Gravitas/AutoGPT/issues/{N}/comments -f body="🤖 Fixed in [abc1234](https://github.com/Significant-Gravitas/AutoGPT/commit/FULL_SHA): <description>"` |
-
-## Codecov coverage
-
-Codecov patch target is **80%** on changed lines. Checks are **informational** (not blocking) but should be green.
-
-### Running coverage locally
-
-**Backend** (from `autogpt_platform/backend/`):
-```bash
-poetry run pytest -s -vv --cov=backend --cov-branch --cov-report term-missing
-```
-
-**Frontend** (from `autogpt_platform/frontend/`):
-```bash
-pnpm vitest run --coverage
-```
-
-### When codecov/patch fails
-
-1. Find uncovered files: `git diff --name-only $(gh pr view --json baseRefName --jq '.baseRefName')...HEAD`
-2. For each uncovered file — extract inline logic to `helpers.ts`/`helpers.py` and test those (highest ROI). Colocate tests as `*_test.py` (backend) or `__tests__/*.test.ts` (frontend).
-3. Run coverage locally to verify, commit, push.
+| Inline review (`pulls/{N}/comments`) | `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/comments/{ID}/replies -f body="🤖 Fixed in <commit-sha>: <description>"` |
+| Conversation (`issues/{N}/comments`) | `gh api repos/Significant-Gravitas/AutoGPT/issues/{N}/comments -f body="🤖 Fixed in <commit-sha>: <description>"` |
 
 ## Format and commit
 

.claude/skills/pr-test/SKILL.md

@@ -547,8 +547,6 @@ Upload screenshots to the PR using the GitHub Git API (no local git operations
 
 **This step is MANDATORY. Every test run MUST post a PR comment with screenshots. No exceptions.**
 
-**CRITICAL — NEVER post a bare directory link like `https://github.com/.../tree/...`.** Every screenshot MUST appear as `![name](raw_url)` inline in the PR comment so reviewers can see them without clicking any links. After posting, the verification step below greps the comment for `![` tags and exits 1 if none are found — the test run is considered incomplete until this passes.
-
 ```bash
 # Upload screenshots via GitHub Git API (creates blobs, tree, commit, and ref remotely)
 REPO="Significant-Gravitas/AutoGPT"
@@ -586,27 +584,15 @@ TREE_JSON+=']'
 
 # Step 2: Create tree, commit, and branch ref
 TREE_SHA=$(echo "$TREE_JSON" | jq -c '{tree: .}' | gh api "repos/${REPO}/git/trees" --input - --jq '.sha')
-
-# Resolve parent commit so screenshots are chained, not orphan root commits
-PARENT_SHA=$(gh api "repos/${REPO}/git/refs/heads/${SCREENSHOTS_BRANCH}" --jq '.object.sha' 2>/dev/null || echo "")
-if [ -n "$PARENT_SHA" ]; then
-  COMMIT_SHA=$(gh api "repos/${REPO}/git/commits" \
-    -f message="test: add E2E test screenshots for PR #${PR_NUMBER}" \
-    -f tree="$TREE_SHA" \
-    -f "parents[]=$PARENT_SHA" \
-    --jq '.sha')
-else
-  COMMIT_SHA=$(gh api "repos/${REPO}/git/commits" \
-    -f message="test: add E2E test screenshots for PR #${PR_NUMBER}" \
-    -f tree="$TREE_SHA" \
-    --jq '.sha')
-fi
-
+COMMIT_SHA=$(gh api "repos/${REPO}/git/commits" \
+  -f message="test: add E2E test screenshots for PR #${PR_NUMBER}" \
+  -f tree="$TREE_SHA" \
+  --jq '.sha')
 gh api "repos/${REPO}/git/refs" \
   -f ref="refs/heads/${SCREENSHOTS_BRANCH}" \
   -f sha="$COMMIT_SHA" 2>/dev/null \
   || gh api "repos/${REPO}/git/refs/heads/${SCREENSHOTS_BRANCH}" \
-    -X PATCH -f sha="$COMMIT_SHA" -F force=true
+    -X PATCH -f sha="$COMMIT_SHA" -f force=true
 ```
 
 Then post the comment with **inline images AND explanations for each screenshot**:
@@ -672,15 +658,6 @@ INNEREOF
 
 gh api "repos/${REPO}/issues/$PR_NUMBER/comments" -F body=@"$COMMENT_FILE"
 rm -f "$COMMENT_FILE"
-
-# Verify the posted comment contains inline images — exit 1 if none found
-# Use separate --paginate + jq pipe: --jq applies per-page, not to the full list
-LAST_COMMENT=$(gh api "repos/${REPO}/issues/$PR_NUMBER/comments" --paginate 2>/dev/null | jq -r '.[-1].body // ""')
-if ! echo "$LAST_COMMENT" | grep -q '!\['; then
-  echo "ERROR: Posted comment contains no inline images (![). Bare directory links are not acceptable." >&2
-  exit 1
-fi
-echo "✓ Inline images verified in posted comment"
 ```
 
 **The PR comment MUST include:**
@@ -690,103 +667,6 @@ echo "✓ Inline images verified in posted comment"
 
 This approach uses the GitHub Git API to create blobs, trees, commits, and refs entirely server-side. No local `git checkout` or `git push` — safe for worktrees and won't interfere with the PR branch.
 
-## Step 8: Evaluate and post a formal PR review
-
-After the test comment is posted, evaluate whether the run was thorough enough to make a merge decision, then post a formal GitHub review (approve or request changes). **This step is mandatory — every test run MUST end with a formal review decision.**
-
-### Evaluation criteria
-
-Re-read the PR description:
-```bash
-gh pr view "$PR_NUMBER" --json body --jq '.body' --repo "$REPO"
-```
-
-Score the run against each criterion:
-
-| Criterion | Pass condition |
-|-----------|---------------|
-| **Coverage** | Every feature/change described in the PR has at least one test scenario |
-| **All scenarios pass** | No FAIL rows in the results table |
-| **Negative tests** | At least one failure-path test per feature (invalid input, unauthorized, edge case) |
-| **Before/after evidence** | Every state-changing API call has before/after values logged |
-| **Screenshots are meaningful** | Screenshots show the actual state change, not just a loading spinner or blank page |
-| **No regressions** | Existing core flows (login, agent create/run) still work |
-
-### Decision logic
-
-```
-ALL criteria pass                            → APPROVE
-Any scenario FAIL or missing PR feature      → REQUEST_CHANGES (list gaps)
-Evidence weak (no before/after, vague shots) → REQUEST_CHANGES (list what's missing)
-```
-
-### Post the review
-
-```bash
-REVIEW_FILE=$(mktemp)
-
-# Count results
-PASS_COUNT=$(echo "$TEST_RESULTS_TABLE" | grep -c "PASS" || true)
-FAIL_COUNT=$(echo "$TEST_RESULTS_TABLE" | grep -c "FAIL" || true)
-TOTAL=$(( PASS_COUNT + FAIL_COUNT ))
-
-# List any coverage gaps found during evaluation (populate this array as you assess)
-# e.g. COVERAGE_GAPS=("PR claims to add X but no test covers it")
-COVERAGE_GAPS=()
-```
-
-**If APPROVING** — all criteria met, zero failures, full coverage:
-
-```bash
-cat > "$REVIEW_FILE" <<REVIEWEOF
-## E2E Test Evaluation — APPROVED
-
-**Results:** ${PASS_COUNT}/${TOTAL} scenarios passed.
-
-**Coverage:** All features described in the PR were exercised.
-
-**Evidence:** Before/after API values logged for all state-changing operations; screenshots show meaningful state transitions.
-
-**Negative tests:** Failure paths tested for each feature.
-
-No regressions observed on core flows.
-REVIEWEOF
-
-gh pr review "$PR_NUMBER" --repo "$REPO" --approve --body "$(cat "$REVIEW_FILE")"
-echo "✅ PR approved"
-```
-
-**If REQUESTING CHANGES** — any failure, coverage gap, or missing evidence:
-
-```bash
-FAIL_LIST=$(echo "$TEST_RESULTS_TABLE" | grep "FAIL" | awk -F'|' '{print "- Scenario" $2 "failed"}' || true)
-
-cat > "$REVIEW_FILE" <<REVIEWEOF
-## E2E Test Evaluation — Changes Requested
-
-**Results:** ${PASS_COUNT}/${TOTAL} scenarios passed, ${FAIL_COUNT} failed.
-
-### Required before merge
-
-${FAIL_LIST}
-$(for gap in "${COVERAGE_GAPS[@]}"; do echo "- $gap"; done)
-
-Please fix the above and re-run the E2E tests.
-REVIEWEOF
-
-gh pr review "$PR_NUMBER" --repo "$REPO" --request-changes --body "$(cat "$REVIEW_FILE")"
-echo "❌ Changes requested"
-```
-
-```bash
-rm -f "$REVIEW_FILE"
-```
-
-**Rules:**
-- In `--fix` mode, fix all failures before posting the review — the review reflects the final state after fixes
-- Never approve if any scenario failed, even if it seems like a flake — rerun that scenario first
-- Never request changes for issues already fixed in this run
-
 ## Fix mode (--fix flag)
 
 When `--fix` is present, the standard is HIGHER. Do not just note issues — FIX them immediately.

.claude/skills/write-frontend-tests/SKILL.md

@@ -1,224 +0,0 @@
----
-name: write-frontend-tests
-description: "Analyze the current branch diff against dev, plan integration tests for changed frontend pages/components, and write them. TRIGGER when user asks to write frontend tests, add test coverage, or 'write tests for my changes'."
-user-invocable: true
-args: "[base branch] — defaults to dev. Optionally pass a specific base branch to diff against."
-metadata:
-  author: autogpt-team
-  version: "1.0.0"
----
-
-# Write Frontend Tests
-
-Analyze the current branch's frontend changes, plan integration tests, and write them.
-
-## References
-
-Before writing any tests, read the testing rules and conventions:
-
-- `autogpt_platform/frontend/TESTING.md` — testing strategy, file locations, examples
-- `autogpt_platform/frontend/src/tests/AGENTS.md` — detailed testing rules, MSW patterns, decision flowchart
-- `autogpt_platform/frontend/src/tests/integrations/test-utils.tsx` — custom render with providers
-- `autogpt_platform/frontend/src/tests/integrations/vitest.setup.tsx` — MSW server setup
-
-## Step 1: Identify changed frontend files
-
-```bash
-BASE_BRANCH="${ARGUMENTS:-dev}"
-cd autogpt_platform/frontend
-
-# Get changed frontend files (excluding generated, config, and test files)
-git diff "$BASE_BRANCH"...HEAD --name-only -- src/ \
-  | grep -v '__generated__' \
-  | grep -v '__tests__' \
-  | grep -v '\.test\.' \
-  | grep -v '\.stories\.' \
-  | grep -v '\.spec\.'
-```
-
-Also read the diff to understand what changed:
-
-```bash
-git diff "$BASE_BRANCH"...HEAD --stat -- src/
-git diff "$BASE_BRANCH"...HEAD -- src/ | head -500
-```
-
-## Step 2: Categorize changes and find test targets
-
-For each changed file, determine:
-
-1. **Is it a page?** (`page.tsx`) — these are the primary test targets
-2. **Is it a hook?** (`use*.ts`) — test via the page that uses it
-3. **Is it a component?** (`.tsx` in `components/`) — test via the parent page unless it's complex enough to warrant isolation
-4. **Is it a helper?** (`helpers.ts`, `utils.ts`) — unit test directly if pure logic
-
-**Priority order:**
-1. Pages with new/changed data fetching or user interactions
-2. Components with complex internal logic (modals, forms, wizards)
-3. Hooks with non-trivial business logic
-4. Pure helper functions
-
-Skip: styling-only changes, type-only changes, config changes.
-
-## Step 3: Check for existing tests
-
-For each test target, check if tests already exist:
-
-```bash
-# For a page at src/app/(platform)/library/page.tsx
-ls src/app/\(platform\)/library/__tests__/ 2>/dev/null
-
-# For a component at src/app/(platform)/library/components/AgentCard/AgentCard.tsx
-ls src/app/\(platform\)/library/components/AgentCard/__tests__/ 2>/dev/null
-```
-
-Note which targets have no tests (need new files) vs which have tests that need updating.
-
-## Step 4: Identify API endpoints used
-
-For each test target, find which API hooks are used:
-
-```bash
-# Find generated API hook imports in the changed files
-grep -rn 'from.*__generated__/endpoints' src/app/\(platform\)/library/
-grep -rn 'use[A-Z].*V[12]' src/app/\(platform\)/library/
-```
-
-For each API hook found, locate the corresponding MSW handler:
-
-```bash
-# If the page uses useGetV2ListLibraryAgents, find its MSW handlers
-grep -rn 'getGetV2ListLibraryAgents.*Handler' src/app/api/__generated__/endpoints/library/library.msw.ts
-```
-
-List every MSW handler you will need (200 for happy path, 4xx for error paths).
-
-## Step 5: Write the test plan
-
-Before writing code, output a plan as a numbered list:
-
-```
-Test plan for [branch name]:
-
-1. src/app/(platform)/library/__tests__/main.test.tsx (NEW)
-   - Renders page with agent list (MSW 200)
-   - Shows loading state
-   - Shows error state (MSW 422)
-   - Handles empty agent list
-
-2. src/app/(platform)/library/__tests__/search.test.tsx (NEW)
-   - Filters agents by search query
-   - Shows no results message
-   - Clears search
-
-3. src/app/(platform)/library/components/AgentCard/__tests__/AgentCard.test.tsx (UPDATE)
-   - Add test for new "duplicate" action
-```
-
-Present this plan to the user. Wait for confirmation before proceeding. If the user has feedback, adjust the plan.
-
-## Step 6: Write the tests
-
-For each test file in the plan, follow these conventions:
-
-### File structure
-
-```tsx
-import { render, screen, waitFor } from "@/tests/integrations/test-utils";
-import { server } from "@/mocks/mock-server";
-// Import MSW handlers for endpoints the page uses
-import {
-  getGetV2ListLibraryAgentsMockHandler200,
-  getGetV2ListLibraryAgentsMockHandler422,
-} from "@/app/api/__generated__/endpoints/library/library.msw";
-// Import the component under test
-import LibraryPage from "../page";
-
-describe("LibraryPage", () => {
-  test("renders agent list from API", async () => {
-    server.use(getGetV2ListLibraryAgentsMockHandler200());
-
-    render(<LibraryPage />);
-
-    expect(await screen.findByText(/my agents/i)).toBeDefined();
-  });
-
-  test("shows error state on API failure", async () => {
-    server.use(getGetV2ListLibraryAgentsMockHandler422());
-
-    render(<LibraryPage />);
-
-    expect(await screen.findByText(/error/i)).toBeDefined();
-  });
-});
-```
-
-### Rules
-
-- Use `render()` from `@/tests/integrations/test-utils` (NOT from `@testing-library/react` directly)
-- Use `server.use()` to set up MSW handlers BEFORE rendering
-- Use `findBy*` (async) for elements that appear after data fetching — NOT `getBy*`
-- Use `getBy*` only for elements that are immediately present in the DOM
-- Use `screen` queries — do NOT destructure from `render()`
-- Use `waitFor` when asserting side effects or state changes after interactions
-- Import `fireEvent` or `userEvent` from the test-utils for interactions
-- Do NOT mock internal hooks or functions — mock at the API boundary via MSW
-- Do NOT use `act()` manually — `render` and `fireEvent` handle it
-- Keep tests focused: one behavior per test
-- Use descriptive test names that read like sentences
-
-### Test location
-
-```
-# For pages: __tests__/ next to page.tsx
-src/app/(platform)/library/__tests__/main.test.tsx
-
-# For complex standalone components: __tests__/ inside component folder
-src/app/(platform)/library/components/AgentCard/__tests__/AgentCard.test.tsx
-
-# For pure helpers: co-located .test.ts
-src/app/(platform)/library/helpers.test.ts
-```
-
-### Custom MSW overrides
-
-When the auto-generated faker data is not enough, override with specific data:
-
-```tsx
-import { http, HttpResponse } from "msw";
-
-server.use(
-  http.get("http://localhost:3000/api/proxy/api/v2/library/agents", () => {
-    return HttpResponse.json({
-      agents: [
-        { id: "1", name: "Test Agent", description: "A test agent" },
-      ],
-      pagination: { total_items: 1, total_pages: 1, page: 1, page_size: 10 },
-    });
-  }),
-);
-```
-
-Use the proxy URL pattern: `http://localhost:3000/api/proxy/api/v{version}/{path}` — this matches the MSW base URL configured in `orval.config.ts`.
-
-## Step 7: Run and verify
-
-After writing all tests:
-
-```bash
-cd autogpt_platform/frontend
-pnpm test:unit --reporter=verbose
-```
-
-If tests fail:
-1. Read the error output carefully
-2. Fix the test (not the source code, unless there is a genuine bug)
-3. Re-run until all pass
-
-Then run the full checks:
-
-```bash
-pnpm format
-pnpm lint
-pnpm types
-```

.github/workflows/classic-autogpt-ci.yml

@@ -6,19 +6,11 @@ on:
     paths:
       - '.github/workflows/classic-autogpt-ci.yml'
       - 'classic/original_autogpt/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/forge/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
   pull_request:
     branches: [ master, dev, release-* ]
     paths:
       - '.github/workflows/classic-autogpt-ci.yml'
       - 'classic/original_autogpt/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/forge/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
 
 concurrency:
   group: ${{ format('classic-autogpt-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
@@ -27,22 +19,47 @@ concurrency:
 defaults:
   run:
     shell: bash
-    working-directory: classic
+    working-directory: classic/original_autogpt
 
 jobs:
   test:
     permissions:
       contents: read
     timeout-minutes: 30
-    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10"]
+        platform-os: [ubuntu, macos, macos-arm64, windows]
+    runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
 
     steps:
-      - name: Start MinIO service
+      # Quite slow on macOS (2~4 minutes to set up Docker)
+      # - name: Set up Docker (macOS)
+      #   if: runner.os == 'macOS'
+      #   uses: crazy-max/ghaction-setup-docker@v3
+
+      - name: Start MinIO service (Linux)
+        if: runner.os == 'Linux'
         working-directory: '.'
         run: |
           docker pull minio/minio:edge-cicd
           docker run -d -p 9000:9000 minio/minio:edge-cicd
 
+      - name: Start MinIO service (macOS)
+        if: runner.os == 'macOS'
+        working-directory: ${{ runner.temp }}
+        run: |
+          brew install minio/stable/minio
+          mkdir data
+          minio server ./data &
+
+      # No MinIO on Windows:
+      # - Windows doesn't support running Linux Docker containers
+      # - It doesn't seem possible to start background processes on Windows. They are
+      #   killed after the step returns.
+      #   See: https://github.com/actions/runner/issues/598#issuecomment-2011890429
+
       - name: Checkout repository
         uses: actions/checkout@v4
         with:
@@ -54,23 +71,41 @@ jobs:
           git config --global user.name "Auto-GPT-Bot"
           git config --global user.email "github-bot@agpt.co"
 
-      - name: Set up Python 3.12
+      - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
-          python-version: "3.12"
+          python-version: ${{ matrix.python-version }}
 
       - id: get_date
         name: Get date
         run: echo "date=$(date +'%Y-%m-%d')" >> $GITHUB_OUTPUT
 
       - name: Set up Python dependency cache
+        # On Windows, unpacking cached dependencies takes longer than just installing them
+        if: runner.os != 'Windows'
         uses: actions/cache@v4
         with:
-          path: ~/.cache/pypoetry
-          key: poetry-${{ runner.os }}-${{ hashFiles('classic/poetry.lock') }}
+          path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
+          key: poetry-${{ runner.os }}-${{ hashFiles('classic/original_autogpt/poetry.lock') }}
 
-      - name: Install Poetry
-        run: curl -sSL https://install.python-poetry.org | python3 -
+      - name: Install Poetry (Unix)
+        if: runner.os != 'Windows'
+        run: |
+          curl -sSL https://install.python-poetry.org | python3 -
+
+          if [ "${{ runner.os }}" = "macOS" ]; then
+            PATH="$HOME/.local/bin:$PATH"
+            echo "$HOME/.local/bin" >> $GITHUB_PATH
+          fi
+
+      - name: Install Poetry (Windows)
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          (Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
+
+          $env:PATH += ";$env:APPDATA\Python\Scripts"
+          echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
 
       - name: Install Python dependencies
         run: poetry install
@@ -81,13 +116,12 @@ jobs:
             --cov=autogpt --cov-branch --cov-report term-missing --cov-report xml \
             --numprocesses=logical --durations=10 \
             --junitxml=junit.xml -o junit_family=legacy \
-            original_autogpt/tests/unit original_autogpt/tests/integration
+            tests/unit tests/integration
         env:
           CI: true
           PLAIN_OUTPUT: True
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-          S3_ENDPOINT_URL: http://127.0.0.1:9000
+          S3_ENDPOINT_URL: ${{ runner.os != 'Windows' && 'http://127.0.0.1:9000' || '' }}
           AWS_ACCESS_KEY_ID: minioadmin
           AWS_SECRET_ACCESS_KEY: minioadmin
 
@@ -101,11 +135,11 @@ jobs:
         uses: codecov/codecov-action@v5
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
-          flags: autogpt-agent
+          flags: autogpt-agent,${{ runner.os }}
 
       - name: Upload logs to artifact
         if: always()
         uses: actions/upload-artifact@v4
         with:
           name: test-logs
-          path: classic/logs/
+          path: classic/original_autogpt/logs/

.github/workflows/classic-autogpt-docker-ci.yml

@@ -148,7 +148,7 @@ jobs:
             --entrypoint poetry ${{ env.IMAGE_NAME }} run \
             pytest -v --cov=autogpt --cov-branch --cov-report term-missing \
             --numprocesses=4 --durations=10 \
-            original_autogpt/tests/unit original_autogpt/tests/integration 2>&1 | tee test_output.txt
+            tests/unit tests/integration 2>&1 | tee test_output.txt
 
           test_failure=${PIPESTATUS[0]}
 

.github/workflows/classic-autogpts-ci.yml

@@ -10,9 +10,10 @@ on:
       - '.github/workflows/classic-autogpts-ci.yml'
       - 'classic/original_autogpt/**'
       - 'classic/forge/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - 'classic/benchmark/**'
+      - 'classic/run'
+      - 'classic/cli.py'
+      - 'classic/setup.py'
       - '!**/*.md'
   pull_request:
     branches: [ master, dev, release-* ]
@@ -20,9 +21,10 @@ on:
       - '.github/workflows/classic-autogpts-ci.yml'
       - 'classic/original_autogpt/**'
       - 'classic/forge/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - 'classic/benchmark/**'
+      - 'classic/run'
+      - 'classic/cli.py'
+      - 'classic/setup.py'
       - '!**/*.md'
 
 defaults:
@@ -33,9 +35,13 @@ defaults:
 jobs:
   serve-agent-protocol:
     runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        agent-name: [ original_autogpt ]
+      fail-fast: false
     timeout-minutes: 20
     env:
-      min-python-version: '3.12'
+      min-python-version: '3.10'
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
@@ -49,22 +55,22 @@ jobs:
           python-version: ${{ env.min-python-version }}
 
       - name: Install Poetry
+        working-directory: ./classic/${{ matrix.agent-name }}/
         run: |
           curl -sSL https://install.python-poetry.org | python -
 
-      - name: Install dependencies
-        run: poetry install
-
-      - name: Run smoke tests with direct-benchmark
+      - name: Run regression tests
         run: |
-          poetry run direct-benchmark run \
-            --strategies one_shot \
-            --models claude \
-            --tests ReadFile,WriteFile \
-            --json
+          ./run agent start ${{ matrix.agent-name }}
+          cd ${{ matrix.agent-name }}
+          poetry run agbenchmark --mock --test=BasicRetrieval --test=Battleship --test=WebArenaTask_0
+          poetry run agbenchmark --test=WriteFile
         env:
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+          AGENT_NAME: ${{ matrix.agent-name }}
           REQUESTS_CA_BUNDLE: /etc/ssl/certs/ca-certificates.crt
-          NONINTERACTIVE_MODE: "true"
-          CI: true
+          HELICONE_CACHE_ENABLED: false
+          HELICONE_PROPERTY_AGENT: ${{ matrix.agent-name }}
+          REPORTS_FOLDER: ${{ format('../../reports/{0}', matrix.agent-name) }}
+          TELEMETRY_ENVIRONMENT: autogpt-ci
+          TELEMETRY_OPT_IN: ${{ github.ref_name == 'master' }}

.github/workflows/classic-benchmark-ci.yml

@@ -1,24 +1,18 @@
-name: Classic - Direct Benchmark CI
+name: Classic - AGBenchmark CI
 
 on:
   push:
     branches: [ master, dev, ci-test* ]
     paths:
-      - 'classic/direct_benchmark/**'
-      - 'classic/original_autogpt/**'
-      - 'classic/forge/**'
+      - 'classic/benchmark/**'
+      - '!classic/benchmark/reports/**'
       - .github/workflows/classic-benchmark-ci.yml
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
   pull_request:
     branches: [ master, dev, release-* ]
     paths:
-      - 'classic/direct_benchmark/**'
-      - 'classic/original_autogpt/**'
-      - 'classic/forge/**'
+      - 'classic/benchmark/**'
+      - '!classic/benchmark/reports/**'
       - .github/workflows/classic-benchmark-ci.yml
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
 
 concurrency:
   group: ${{ format('benchmark-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
@@ -29,16 +23,23 @@ defaults:
     shell: bash
 
 env:
-  min-python-version: '3.12'
+  min-python-version: '3.10'
 
 jobs:
-  benchmark-tests:
-    runs-on: ubuntu-latest
+  test:
+    permissions:
+      contents: read
     timeout-minutes: 30
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10"]
+        platform-os: [ubuntu, macos, macos-arm64, windows]
+    runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
     defaults:
       run:
         shell: bash
-        working-directory: classic
+        working-directory: classic/benchmark
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
@@ -46,88 +47,71 @@ jobs:
           fetch-depth: 0
           submodules: true
 
-      - name: Set up Python ${{ env.min-python-version }}
+      - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
-          python-version: ${{ env.min-python-version }}
+          python-version: ${{ matrix.python-version }}
 
       - name: Set up Python dependency cache
+        # On Windows, unpacking cached dependencies takes longer than just installing them
+        if: runner.os != 'Windows'
         uses: actions/cache@v4
         with:
-          path: ~/.cache/pypoetry
-          key: poetry-${{ runner.os }}-${{ hashFiles('classic/poetry.lock') }}
+          path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
+          key: poetry-${{ runner.os }}-${{ hashFiles('classic/benchmark/poetry.lock') }}
 
-      - name: Install Poetry
+      - name: Install Poetry (Unix)
+        if: runner.os != 'Windows'
         run: |
           curl -sSL https://install.python-poetry.org | python3 -
 
-      - name: Install dependencies
+          if [ "${{ runner.os }}" = "macOS" ]; then
+            PATH="$HOME/.local/bin:$PATH"
+            echo "$HOME/.local/bin" >> $GITHUB_PATH
+          fi
+
+      - name: Install Poetry (Windows)
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          (Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
+
+          $env:PATH += ";$env:APPDATA\Python\Scripts"
+          echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
+
+      - name: Install Python dependencies
         run: poetry install
 
-      - name: Run basic benchmark tests
+      - name: Run pytest with coverage
         run: |
-          echo "Testing ReadFile challenge with one_shot strategy..."
-          poetry run direct-benchmark run \
-            --fresh \
-            --strategies one_shot \
-            --models claude \
-            --tests ReadFile \
-            --json
-
-          echo "Testing WriteFile challenge..."
-          poetry run direct-benchmark run \
-            --fresh \
-            --strategies one_shot \
-            --models claude \
-            --tests WriteFile \
-            --json
+          poetry run pytest -vv \
+            --cov=agbenchmark --cov-branch --cov-report term-missing --cov-report xml \
+            --durations=10 \
+            --junitxml=junit.xml -o junit_family=legacy \
+            tests
         env:
           CI: true
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          NONINTERACTIVE_MODE: "true"
 
-      - name: Test category filtering
-        run: |
-          echo "Testing coding category..."
-          poetry run direct-benchmark run \
-            --fresh \
-            --strategies one_shot \
-            --models claude \
-            --categories coding \
-            --tests ReadFile,WriteFile \
-            --json
-        env:
-          CI: true
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          NONINTERACTIVE_MODE: "true"
+      - name: Upload test results to Codecov
+        if: ${{ !cancelled() }}  # Run even if tests fail
+        uses: codecov/test-results-action@v1
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
 
-      - name: Test multiple strategies
-        run: |
-          echo "Testing multiple strategies..."
-          poetry run direct-benchmark run \
-            --fresh \
-            --strategies one_shot,plan_execute \
-            --models claude \
-            --tests ReadFile \
-            --parallel 2 \
-            --json
-        env:
-          CI: true
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          NONINTERACTIVE_MODE: "true"
+      - name: Upload coverage reports to Codecov
+        uses: codecov/codecov-action@v5
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          flags: agbenchmark,${{ runner.os }}
 
-  # Run regression tests on maintain challenges
-  regression-tests:
+  self-test-with-agent:
     runs-on: ubuntu-latest
-    timeout-minutes: 45
-    if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/dev'
-    defaults:
-      run:
-        shell: bash
-        working-directory: classic
+    strategy:
+      matrix:
+        agent-name: [forge]
+      fail-fast: false
+    timeout-minutes: 20
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
@@ -140,31 +124,53 @@ jobs:
         with:
           python-version: ${{ env.min-python-version }}
 
-      - name: Set up Python dependency cache
-        uses: actions/cache@v4
-        with:
-          path: ~/.cache/pypoetry
-          key: poetry-${{ runner.os }}-${{ hashFiles('classic/poetry.lock') }}
-
       - name: Install Poetry
         run: |
-          curl -sSL https://install.python-poetry.org | python3 -
-
-      - name: Install dependencies
-        run: poetry install
+          curl -sSL https://install.python-poetry.org | python -
 
       - name: Run regression tests
+        working-directory: classic
         run: |
-          echo "Running regression tests (previously beaten challenges)..."
-          poetry run direct-benchmark run \
-            --fresh \
-            --strategies one_shot \
-            --models claude \
-            --maintain \
-            --parallel 4 \
-            --json
+          ./run agent start ${{ matrix.agent-name }}
+          cd ${{ matrix.agent-name }}
+
+          set +e # Ignore non-zero exit codes and continue execution
+          echo "Running the following command: poetry run agbenchmark --maintain --mock"
+          poetry run agbenchmark --maintain --mock
+          EXIT_CODE=$?
+          set -e  # Stop ignoring non-zero exit codes
+          # Check if the exit code was 5, and if so, exit with 0 instead
+          if [ $EXIT_CODE -eq 5 ]; then
+            echo "regression_tests.json is empty."
+          fi
+
+          echo "Running the following command: poetry run agbenchmark --mock"
+          poetry run agbenchmark --mock
+
+          echo "Running the following command: poetry run agbenchmark --mock --category=data"
+          poetry run agbenchmark --mock --category=data
+
+          echo "Running the following command: poetry run agbenchmark --mock --category=coding"
+          poetry run agbenchmark --mock --category=coding
+
+          # echo "Running the following command: poetry run agbenchmark --test=WriteFile"
+          # poetry run agbenchmark --test=WriteFile
+          cd ../benchmark
+          poetry install
+          echo "Adding the BUILD_SKILL_TREE environment variable. This will attempt to add new elements in the skill tree. If new elements are added, the CI fails because they should have been pushed"
+          export BUILD_SKILL_TREE=true
+
+          # poetry run agbenchmark --mock
+
+          # CHANGED=$(git diff --name-only | grep -E '(agbenchmark/challenges)|(../classic/frontend/assets)') || echo "No diffs"
+          # if [ ! -z "$CHANGED" ]; then
+          #   echo "There are unstaged changes please run agbenchmark and commit those changes since they are needed."
+          #   echo "$CHANGED"
+          #   exit 1
+          # else
+          #   echo "No unstaged changes."
+          # fi
         env:
-          CI: true
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          NONINTERACTIVE_MODE: "true"
+          TELEMETRY_ENVIRONMENT: autogpt-benchmark-ci
+          TELEMETRY_OPT_IN: ${{ github.ref_name == 'master' }}

.github/workflows/classic-forge-ci.yml

@@ -6,15 +6,13 @@ on:
     paths:
       - '.github/workflows/classic-forge-ci.yml'
       - 'classic/forge/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - '!classic/forge/tests/vcr_cassettes'
   pull_request:
     branches: [ master, dev, release-* ]
     paths:
       - '.github/workflows/classic-forge-ci.yml'
       - 'classic/forge/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - '!classic/forge/tests/vcr_cassettes'
 
 concurrency:
   group: ${{ format('forge-ci-{0}', github.head_ref && format('{0}-{1}', github.event_name, github.event.pull_request.number) || github.sha) }}
@@ -23,60 +21,131 @@ concurrency:
 defaults:
   run:
     shell: bash
-    working-directory: classic
+    working-directory: classic/forge
 
 jobs:
   test:
     permissions:
       contents: read
     timeout-minutes: 30
-    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10"]
+        platform-os: [ubuntu, macos, macos-arm64, windows]
+    runs-on: ${{ matrix.platform-os != 'macos-arm64' && format('{0}-latest', matrix.platform-os) || 'macos-14' }}
 
     steps:
-      - name: Start MinIO service
+      # Quite slow on macOS (2~4 minutes to set up Docker)
+      # - name: Set up Docker (macOS)
+      #   if: runner.os == 'macOS'
+      #   uses: crazy-max/ghaction-setup-docker@v3
+
+      - name: Start MinIO service (Linux)
+        if: runner.os == 'Linux'
         working-directory: '.'
         run: |
           docker pull minio/minio:edge-cicd
           docker run -d -p 9000:9000 minio/minio:edge-cicd
 
+      - name: Start MinIO service (macOS)
+        if: runner.os == 'macOS'
+        working-directory: ${{ runner.temp }}
+        run: |
+          brew install minio/stable/minio
+          mkdir data
+          minio server ./data &
+
+      # No MinIO on Windows:
+      # - Windows doesn't support running Linux Docker containers
+      # - It doesn't seem possible to start background processes on Windows. They are
+      #   killed after the step returns.
+      #   See: https://github.com/actions/runner/issues/598#issuecomment-2011890429
+
       - name: Checkout repository
         uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          submodules: true
 
-      - name: Set up Python 3.12
+      - name: Checkout cassettes
+        if: ${{ startsWith(github.event_name, 'pull_request') }}
+        env:
+          PR_BASE: ${{ github.event.pull_request.base.ref }}
+          PR_BRANCH: ${{ github.event.pull_request.head.ref }}
+          PR_AUTHOR: ${{ github.event.pull_request.user.login }}
+        run: |
+          cassette_branch="${PR_AUTHOR}-${PR_BRANCH}"
+          cassette_base_branch="${PR_BASE}"
+          cd tests/vcr_cassettes
+
+          if ! git ls-remote --exit-code --heads origin $cassette_base_branch ; then
+            cassette_base_branch="master"
+          fi
+
+          if git ls-remote --exit-code --heads origin $cassette_branch ; then
+            git fetch origin $cassette_branch
+            git fetch origin $cassette_base_branch
+
+            git checkout $cassette_branch
+
+            # Pick non-conflicting cassette updates from the base branch
+            git merge --no-commit --strategy-option=ours origin/$cassette_base_branch
+            echo "Using cassettes from mirror branch '$cassette_branch'," \
+              "synced to upstream branch '$cassette_base_branch'."
+          else
+            git checkout -b $cassette_branch
+            echo "Branch '$cassette_branch' does not exist in cassette submodule." \
+              "Using cassettes from '$cassette_base_branch'."
+          fi
+
+      - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v5
         with:
-          python-version: "3.12"
+          python-version: ${{ matrix.python-version }}
 
       - name: Set up Python dependency cache
+        # On Windows, unpacking cached dependencies takes longer than just installing them
+        if: runner.os != 'Windows'
         uses: actions/cache@v4
         with:
-          path: ~/.cache/pypoetry
-          key: poetry-${{ runner.os }}-${{ hashFiles('classic/poetry.lock') }}
+          path: ${{ runner.os == 'macOS' && '~/Library/Caches/pypoetry' || '~/.cache/pypoetry' }}
+          key: poetry-${{ runner.os }}-${{ hashFiles('classic/forge/poetry.lock') }}
 
-      - name: Install Poetry
-        run: curl -sSL https://install.python-poetry.org | python3 -
+      - name: Install Poetry (Unix)
+        if: runner.os != 'Windows'
+        run: |
+          curl -sSL https://install.python-poetry.org | python3 -
+
+          if [ "${{ runner.os }}" = "macOS" ]; then
+            PATH="$HOME/.local/bin:$PATH"
+            echo "$HOME/.local/bin" >> $GITHUB_PATH
+          fi
+
+      - name: Install Poetry (Windows)
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          (Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -
+
+          $env:PATH += ";$env:APPDATA\Python\Scripts"
+          echo "$env:APPDATA\Python\Scripts" >> $env:GITHUB_PATH
 
       - name: Install Python dependencies
         run: poetry install
 
-      - name: Install Playwright browsers
-        run: poetry run playwright install chromium
-
       - name: Run pytest with coverage
         run: |
           poetry run pytest -vv \
             --cov=forge --cov-branch --cov-report term-missing --cov-report xml \
             --durations=10 \
             --junitxml=junit.xml -o junit_family=legacy \
-            forge/forge forge/tests
+            forge
         env:
           CI: true
           PLAIN_OUTPUT: True
-          # API keys - tests that need these will skip if not available
-          # Secrets are not available to fork PRs (GitHub security feature)
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-          S3_ENDPOINT_URL: http://127.0.0.1:9000
+          S3_ENDPOINT_URL: ${{ runner.os != 'Windows' && 'http://127.0.0.1:9000' || '' }}
           AWS_ACCESS_KEY_ID: minioadmin
           AWS_SECRET_ACCESS_KEY: minioadmin
 
@@ -90,11 +159,85 @@ jobs:
         uses: codecov/codecov-action@v5
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
-          flags: forge
+          flags: forge,${{ runner.os }}
+
+      - id: setup_git_auth
+        name: Set up git token authentication
+        # Cassettes may be pushed even when tests fail
+        if: success() || failure()
+        run: |
+          config_key="http.${{ github.server_url }}/.extraheader"
+          if [ "${{ runner.os }}" = 'macOS' ]; then
+            base64_pat=$(echo -n "pat:${{ secrets.PAT_REVIEW }}" | base64)
+          else
+            base64_pat=$(echo -n "pat:${{ secrets.PAT_REVIEW }}" | base64 -w0)
+          fi
+
+          git config "$config_key" \
+            "Authorization: Basic $base64_pat"
+
+          cd tests/vcr_cassettes
+          git config "$config_key" \
+            "Authorization: Basic $base64_pat"
+
+          echo "config_key=$config_key" >> $GITHUB_OUTPUT
+
+      - id: push_cassettes
+        name: Push updated cassettes
+        # For pull requests, push updated cassettes even when tests fail
+        if: github.event_name == 'push' || (! github.event.pull_request.head.repo.fork && (success() || failure()))
+        env:
+          PR_BRANCH: ${{ github.event.pull_request.head.ref }}
+          PR_AUTHOR: ${{ github.event.pull_request.user.login }}
+        run: |
+          if [ "${{ startsWith(github.event_name, 'pull_request') }}" = "true" ]; then
+            is_pull_request=true
+            cassette_branch="${PR_AUTHOR}-${PR_BRANCH}"
+          else
+            cassette_branch="${{ github.ref_name }}"
+          fi
+
+          cd tests/vcr_cassettes
+          # Commit & push changes to cassettes if any
+          if ! git diff --quiet; then
+            git add .
+            git commit -m "Auto-update cassettes"
+            git push origin HEAD:$cassette_branch
+            if [ ! $is_pull_request ]; then
+              cd ../..
+              git add tests/vcr_cassettes
+              git commit -m "Update cassette submodule"
+              git push origin HEAD:$cassette_branch
+            fi
+            echo "updated=true" >> $GITHUB_OUTPUT
+          else
+            echo "updated=false" >> $GITHUB_OUTPUT
+            echo "No cassette changes to commit"
+          fi
+
+      - name: Post Set up git token auth
+        if: steps.setup_git_auth.outcome == 'success'
+        run: |
+          git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
+          git submodule foreach git config --unset-all '${{ steps.setup_git_auth.outputs.config_key }}'
+
+      - name: Apply "behaviour change" label and comment on PR
+        if: ${{ startsWith(github.event_name, 'pull_request') }}
+        run: |
+          PR_NUMBER="${{ github.event.pull_request.number }}"
+          TOKEN="${{ secrets.PAT_REVIEW }}"
+          REPO="${{ github.repository }}"
+
+          if [[ "${{ steps.push_cassettes.outputs.updated }}" == "true" ]]; then
+            echo "Adding label and comment..."
+            echo $TOKEN | gh auth login --with-token
+            gh issue edit $PR_NUMBER --add-label "behaviour change"
+            gh issue comment $PR_NUMBER --body "You changed AutoGPT's behaviour on ${{ runner.os }}. The cassettes have been updated and will be merged to the submodule when this Pull Request gets merged."
+          fi
 
       - name: Upload logs to artifact
         if: always()
         uses: actions/upload-artifact@v4
         with:
           name: test-logs
-          path: classic/logs/
+          path: classic/forge/logs/

.github/workflows/classic-frontend-ci.yml

@@ -0,0 +1,60 @@
+name: Classic - Frontend CI/CD
+
+on:
+  push:
+    branches:
+      - master
+      - dev
+      - 'ci-test*' # This will match any branch that starts with "ci-test"
+    paths:
+      - 'classic/frontend/**'
+      - '.github/workflows/classic-frontend-ci.yml'
+  pull_request:
+    paths:
+      - 'classic/frontend/**'
+      - '.github/workflows/classic-frontend-ci.yml'
+
+jobs:
+  build:
+    permissions:
+      contents: write
+      pull-requests: write
+    runs-on: ubuntu-latest
+    env:
+      BUILD_BRANCH: ${{ format('classic-frontend-build/{0}', github.ref_name) }}
+
+    steps:
+      - name: Checkout Repo
+        uses: actions/checkout@v4
+
+      - name: Setup Flutter
+        uses: subosito/flutter-action@v2
+        with:
+          flutter-version: '3.13.2'
+
+      - name: Build Flutter to Web
+        run: |
+          cd classic/frontend
+          flutter build web --base-href /app/
+
+      # - name: Commit and Push to ${{ env.BUILD_BRANCH }}
+      #   if: github.event_name == 'push'
+      #   run: |
+      #     git config --local user.email "action@github.com"
+      #     git config --local user.name "GitHub Action"
+      #     git add classic/frontend/build/web
+      #     git checkout -B ${{ env.BUILD_BRANCH }}
+      #     git commit -m "Update frontend build to ${GITHUB_SHA:0:7}" -a
+      #     git push -f origin ${{ env.BUILD_BRANCH }}
+
+      - name: Create PR ${{ env.BUILD_BRANCH }} -> ${{ github.ref_name }}
+        if: github.event_name == 'push'
+        uses: peter-evans/create-pull-request@v8
+        with:
+          add-paths: classic/frontend/build/web
+          base: ${{ github.ref_name }}
+          branch: ${{ env.BUILD_BRANCH }}
+          delete-branch: true
+          title: "Update frontend build in `${{ github.ref_name }}`"
+          body: "This PR updates the frontend build based on commit ${{ github.sha }}."
+          commit-message: "Update frontend build based on commit ${{ github.sha }}"

.github/workflows/classic-python-checks.yml

@@ -7,9 +7,7 @@ on:
       - '.github/workflows/classic-python-checks-ci.yml'
       - 'classic/original_autogpt/**'
       - 'classic/forge/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - 'classic/benchmark/**'
       - '**.py'
       - '!classic/forge/tests/vcr_cassettes'
   pull_request:
@@ -18,9 +16,7 @@ on:
       - '.github/workflows/classic-python-checks-ci.yml'
       - 'classic/original_autogpt/**'
       - 'classic/forge/**'
-      - 'classic/direct_benchmark/**'
-      - 'classic/pyproject.toml'
-      - 'classic/poetry.lock'
+      - 'classic/benchmark/**'
       - '**.py'
       - '!classic/forge/tests/vcr_cassettes'
 
@@ -31,13 +27,44 @@ concurrency:
 defaults:
   run:
     shell: bash
-    working-directory: classic
 
 jobs:
+  get-changed-parts:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - id: changes-in
+        name: Determine affected subprojects
+        uses: dorny/paths-filter@v3
+        with:
+          filters: |
+            original_autogpt:
+              - classic/original_autogpt/autogpt/**
+              - classic/original_autogpt/tests/**
+              - classic/original_autogpt/poetry.lock
+            forge:
+              - classic/forge/forge/**
+              - classic/forge/tests/**
+              - classic/forge/poetry.lock
+            benchmark:
+              - classic/benchmark/agbenchmark/**
+              - classic/benchmark/tests/**
+              - classic/benchmark/poetry.lock
+    outputs:
+      changed-parts: ${{ steps.changes-in.outputs.changes }}
+
   lint:
+    needs: get-changed-parts
     runs-on: ubuntu-latest
     env:
-      min-python-version: "3.12"
+      min-python-version: "3.10"
+
+    strategy:
+      matrix:
+        sub-package: ${{ fromJson(needs.get-changed-parts.outputs.changed-parts) }}
+      fail-fast: false
 
     steps:
       - name: Checkout repository
@@ -54,31 +81,42 @@ jobs:
         uses: actions/cache@v4
         with:
           path: ~/.cache/pypoetry
-          key: ${{ runner.os }}-poetry-${{ hashFiles('classic/poetry.lock') }}
+          key: ${{ runner.os }}-poetry-${{ hashFiles(format('{0}/poetry.lock', matrix.sub-package)) }}
 
       - name: Install Poetry
         run: curl -sSL https://install.python-poetry.org | python3 -
 
+      # Install dependencies
+
       - name: Install Python dependencies
-        run: poetry install
+        run: poetry -C classic/${{ matrix.sub-package }} install
 
       # Lint
 
       - name: Lint (isort)
         run: poetry run isort --check .
+        working-directory: classic/${{ matrix.sub-package }}
 
       - name: Lint (Black)
         if: success() || failure()
         run: poetry run black --check .
+        working-directory: classic/${{ matrix.sub-package }}
 
       - name: Lint (Flake8)
         if: success() || failure()
         run: poetry run flake8 .
+        working-directory: classic/${{ matrix.sub-package }}
 
   types:
+    needs: get-changed-parts
     runs-on: ubuntu-latest
     env:
-      min-python-version: "3.12"
+      min-python-version: "3.10"
+
+    strategy:
+      matrix:
+        sub-package: ${{ fromJson(needs.get-changed-parts.outputs.changed-parts) }}
+      fail-fast: false
 
     steps:
       - name: Checkout repository
@@ -95,16 +133,19 @@ jobs:
         uses: actions/cache@v4
         with:
           path: ~/.cache/pypoetry
-          key: ${{ runner.os }}-poetry-${{ hashFiles('classic/poetry.lock') }}
+          key: ${{ runner.os }}-poetry-${{ hashFiles(format('{0}/poetry.lock', matrix.sub-package)) }}
 
       - name: Install Poetry
         run: curl -sSL https://install.python-poetry.org | python3 -
 
+      # Install dependencies
+
       - name: Install Python dependencies
-        run: poetry install
+        run: poetry -C classic/${{ matrix.sub-package }} install
 
       # Typecheck
 
       - name: Typecheck
         if: success() || failure()
         run: poetry run pyright
+        working-directory: classic/${{ matrix.sub-package }}

.github/workflows/platform-backend-ci.yml

@@ -269,14 +269,12 @@ jobs:
           DATABASE_URL: ${{ steps.supabase.outputs.DB_URL }}
           DIRECT_URL: ${{ steps.supabase.outputs.DB_URL }}
 
-      - name: Run pytest with coverage
+      - name: Run pytest
         run: |
           if [[ "${{ runner.debug }}" == "1" ]]; then
-            poetry run pytest -s -vv -o log_cli=true -o log_cli_level=DEBUG \
-              --cov=backend --cov-branch --cov-report term-missing --cov-report xml
+            poetry run pytest -s -vv -o log_cli=true -o log_cli_level=DEBUG
           else
-            poetry run pytest -s -vv \
-              --cov=backend --cov-branch --cov-report term-missing --cov-report xml
+            poetry run pytest -s -vv
           fi
         env:
           LOG_LEVEL: ${{ runner.debug && 'DEBUG' || 'INFO' }}
@@ -289,13 +287,11 @@ jobs:
           REDIS_PORT: "6379"
           ENCRYPTION_KEY: "dvziYgz0KSK8FENhju0ZYi8-fRTfAdlz6YLhdB_jhNw=" # DO NOT USE IN PRODUCTION!!
 
-      - name: Upload coverage reports to Codecov
-        if: ${{ !cancelled() }}
-        uses: codecov/codecov-action@v5
-        with:
-          token: ${{ secrets.CODECOV_TOKEN }}
-          flags: platform-backend
-          files: ./autogpt_platform/backend/coverage.xml
+      # - name: Upload coverage reports to Codecov
+      #   uses: codecov/codecov-action@v4
+      #   with:
+      #     token: ${{ secrets.CODECOV_TOKEN }}
+      #     flags: backend,${{ runner.os }}
 
     env:
       CI: true

.github/workflows/platform-frontend-ci.yml

@@ -148,11 +148,3 @@ jobs:
 
       - name: Run Integration Tests
         run: pnpm test:unit
-
-      - name: Upload coverage reports to Codecov
-        if: ${{ !cancelled() }}
-        uses: codecov/codecov-action@v5
-        with:
-          token: ${{ secrets.CODECOV_TOKEN }}
-          flags: platform-frontend
-          files: ./autogpt_platform/frontend/coverage/cobertura-coverage.xml

.github/workflows/platform-fullstack-ci.yml

@@ -179,30 +179,21 @@ jobs:
           pip install pyyaml
 
           # Resolve extends and generate a flat compose file that bake can understand
-          export NEXT_PUBLIC_SOURCEMAPS NEXT_PUBLIC_PW_TEST
           docker compose -f docker-compose.yml config > docker-compose.resolved.yml
 
-          # Ensure NEXT_PUBLIC_SOURCEMAPS is in resolved compose
-          # (docker compose config on some versions drops this arg)
-          if ! grep -q "NEXT_PUBLIC_SOURCEMAPS" docker-compose.resolved.yml; then
-            echo "Injecting NEXT_PUBLIC_SOURCEMAPS into resolved compose (docker compose config dropped it)"
-            sed -i '/NEXT_PUBLIC_PW_TEST/a\        NEXT_PUBLIC_SOURCEMAPS: "true"' docker-compose.resolved.yml
-          fi
-
           # Add cache configuration to the resolved compose file
           python ../.github/workflows/scripts/docker-ci-fix-compose-build-cache.py \
             --source docker-compose.resolved.yml \
             --cache-from "type=gha" \
             --cache-to "type=gha,mode=max" \
             --backend-hash "${{ hashFiles('autogpt_platform/backend/Dockerfile', 'autogpt_platform/backend/poetry.lock', 'autogpt_platform/backend/backend/**') }}" \
-            --frontend-hash "${{ hashFiles('autogpt_platform/frontend/Dockerfile', 'autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/src/**') }}-sourcemaps" \
+            --frontend-hash "${{ hashFiles('autogpt_platform/frontend/Dockerfile', 'autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/src/**') }}" \
             --git-ref "${{ github.ref }}"
 
           # Build with bake using the resolved compose file (now includes cache config)
           docker buildx bake --allow=fs.read=.. -f docker-compose.resolved.yml --load
         env:
           NEXT_PUBLIC_PW_TEST: true
-          NEXT_PUBLIC_SOURCEMAPS: true
 
       - name: Set up tests - Cache E2E test data
         id: e2e-data-cache
@@ -288,11 +279,6 @@ jobs:
           cache: "pnpm"
           cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
 
-      - name: Copy source maps from Docker for E2E coverage
-        run: |
-          FRONTEND_CONTAINER=$(docker compose -f ../docker-compose.resolved.yml ps -q frontend)
-          docker cp "$FRONTEND_CONTAINER":/app/.next/static .next-static-coverage
-
       - name: Set up tests - Install dependencies
         run: pnpm install --frozen-lockfile
 
@@ -303,15 +289,6 @@ jobs:
         run: pnpm test:no-build
         continue-on-error: false
 
-      - name: Upload E2E coverage to Codecov
-        if: ${{ !cancelled() }}
-        uses: codecov/codecov-action@v5
-        with:
-          token: ${{ secrets.CODECOV_TOKEN }}
-          flags: platform-frontend-e2e
-          files: ./autogpt_platform/frontend/coverage/e2e/cobertura-coverage.xml
-          disable_search: true
-
       - name: Upload Playwright report
         if: always()
         uses: actions/upload-artifact@v4

.gitignore

@@ -3,7 +3,6 @@
 classic/original_autogpt/keys.py
 classic/original_autogpt/*.json
 auto_gpt_workspace/*
-.autogpt/
 *.mpeg
 .env
 # Root .env files
@@ -17,7 +16,6 @@ log-ingestion.txt
 /logs
 *.log
 *.mp3
-!autogpt_platform/frontend/public/notification.mp3
 mem.sqlite3
 venvAutoGPT
 
@@ -161,10 +159,6 @@ CURRENT_BULLETIN.md
 
 # AgBenchmark
 classic/benchmark/agbenchmark/reports/
-classic/reports/
-classic/direct_benchmark/reports/
-classic/.benchmark_workspaces/
-classic/direct_benchmark/.benchmark_workspaces/
 
 # Nodejs
 package-lock.json
@@ -183,13 +177,9 @@ autogpt_platform/backend/settings.py
 
 *.ign.*
 .test-contents
-**/.claude/settings.local.json
 .claude/settings.local.json
 CLAUDE.local.md
 /autogpt_platform/backend/logs
-
-# Test database
-test.db
 .next
 # Implementation plans (generated by AI agents)
 plans/

.gitleaks.toml

@@ -1,36 +0,0 @@
-title = "AutoGPT Gitleaks Config"
-
-[extend]
-useDefault = true
-
-[allowlist]
-description = "Global allowlist"
-paths = [
-    # Template/example env files (no real secrets)
-    '''\.env\.(default|example|template)$''',
-    # Lock files
-    '''pnpm-lock\.yaml$''',
-    '''poetry\.lock$''',
-    # Secrets baseline
-    '''\.secrets\.baseline$''',
-    # Build artifacts and caches (should not be committed)
-    '''__pycache__/''',
-    '''classic/frontend/build/''',
-    # Docker dev setup (local dev JWTs/keys only)
-    '''autogpt_platform/db/docker/''',
-    # Load test configs (dev JWTs)
-    '''load-tests/configs/''',
-    # Test files with fake/fixture keys (_test.py, test_*.py, conftest.py)
-    '''(_test|test_.*|conftest)\.py$''',
-    # Documentation (only contains placeholder keys in curl/API examples)
-    '''docs/.*\.md$''',
-    # Firebase config (public API keys by design)
-    '''google-services\.json$''',
-    '''classic/frontend/(lib|web)/''',
-]
-# CI test-only encryption key (marked DO NOT USE IN PRODUCTION)
-regexes = [
-    '''dvziYgz0KSK8FENhju0ZYi8''',
-    # LLM model name enum values falsely flagged as API keys
-    '''Llama-\d.*Instruct''',
-]

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "classic/forge/tests/vcr_cassettes"]
+	path = classic/forge/tests/vcr_cassettes
+	url = https://github.com/Significant-Gravitas/Auto-GPT-test-cassettes

.pre-commit-config.yaml

@@ -23,15 +23,9 @@ repos:
       - id: detect-secrets
         name: Detect secrets
         description: Detects high entropy strings that are likely to be passwords.
-        args: ["--baseline", ".secrets.baseline"]
         files: ^autogpt_platform/
-        exclude: (pnpm-lock\.yaml|\.env\.(default|example|template))$
-
-  - repo: https://github.com/gitleaks/gitleaks
-    rev: v8.24.3
-    hooks:
-      - id: gitleaks
-        name: Detect secrets (gitleaks)
+        exclude: pnpm-lock\.yaml$
+        stages: [pre-push]
 
   - repo: local
     # For proper type checking, all dependencies need to be up-to-date.
@@ -90,16 +84,51 @@ repos:
         stages: [pre-commit, post-checkout]
 
       - id: poetry-install
-        name: Check & Install dependencies - Classic
-        alias: poetry-install-classic
+        name: Check & Install dependencies - Classic - AutoGPT
+        alias: poetry-install-classic-autogpt
         entry: >
           bash -c '
           if [ -n "$PRE_COMMIT_FROM_REF" ]; then
             git diff --name-only "$PRE_COMMIT_FROM_REF" "$PRE_COMMIT_TO_REF"
           else
             git diff --cached --name-only
-          fi | grep -qE "^classic/poetry\.lock$" || exit 0;
-          poetry -C classic install
+          fi | grep -qE "^classic/(original_autogpt|forge)/poetry\.lock$" || exit 0;
+          poetry -C classic/original_autogpt install
+          '
+        # include forge source (since it's a path dependency)
+        always_run: true
+        language: system
+        pass_filenames: false
+        stages: [pre-commit, post-checkout]
+
+      - id: poetry-install
+        name: Check & Install dependencies - Classic - Forge
+        alias: poetry-install-classic-forge
+        entry: >
+          bash -c '
+          if [ -n "$PRE_COMMIT_FROM_REF" ]; then
+            git diff --name-only "$PRE_COMMIT_FROM_REF" "$PRE_COMMIT_TO_REF"
+          else
+            git diff --cached --name-only
+          fi | grep -qE "^classic/forge/poetry\.lock$" || exit 0;
+          poetry -C classic/forge install
+          '
+        always_run: true
+        language: system
+        pass_filenames: false
+        stages: [pre-commit, post-checkout]
+
+      - id: poetry-install
+        name: Check & Install dependencies - Classic - Benchmark
+        alias: poetry-install-classic-benchmark
+        entry: >
+          bash -c '
+          if [ -n "$PRE_COMMIT_FROM_REF" ]; then
+            git diff --name-only "$PRE_COMMIT_FROM_REF" "$PRE_COMMIT_TO_REF"
+          else
+            git diff --cached --name-only
+          fi | grep -qE "^classic/benchmark/poetry\.lock$" || exit 0;
+          poetry -C classic/benchmark install
           '
         always_run: true
         language: system
@@ -194,10 +223,26 @@ repos:
         language: system
 
       - id: isort
-        name: Lint (isort) - Classic
-        alias: isort-classic
-        entry: bash -c 'cd classic && poetry run isort $(echo "$@" | sed "s|classic/||g")' --
-        files: ^classic/(original_autogpt|forge|direct_benchmark)/
+        name: Lint (isort) - Classic - AutoGPT
+        alias: isort-classic-autogpt
+        entry: poetry -P classic/original_autogpt run isort -p autogpt
+        files: ^classic/original_autogpt/
+        types: [file, python]
+        language: system
+
+      - id: isort
+        name: Lint (isort) - Classic - Forge
+        alias: isort-classic-forge
+        entry: poetry -P classic/forge run isort -p forge
+        files: ^classic/forge/
+        types: [file, python]
+        language: system
+
+      - id: isort
+        name: Lint (isort) - Classic - Benchmark
+        alias: isort-classic-benchmark
+        entry: poetry -P classic/benchmark run isort -p agbenchmark
+        files: ^classic/benchmark/
         types: [file, python]
         language: system
 
@@ -211,13 +256,26 @@ repos:
 
   - repo: https://github.com/PyCQA/flake8
     rev: 7.0.0
-    # Use consolidated flake8 config at classic/.flake8
+    # To have flake8 load the config of the individual subprojects, we have to call
+    # them separately.
     hooks:
       - id: flake8
-        name: Lint (Flake8) - Classic
-        alias: flake8-classic
-        files: ^classic/(original_autogpt|forge|direct_benchmark)/
-        args: [--config=classic/.flake8]
+        name: Lint (Flake8) - Classic - AutoGPT
+        alias: flake8-classic-autogpt
+        files: ^classic/original_autogpt/(autogpt|scripts|tests)/
+        args: [--config=classic/original_autogpt/.flake8]
+
+      - id: flake8
+        name: Lint (Flake8) - Classic - Forge
+        alias: flake8-classic-forge
+        files: ^classic/forge/(forge|tests)/
+        args: [--config=classic/forge/.flake8]
+
+      - id: flake8
+        name: Lint (Flake8) - Classic - Benchmark
+        alias: flake8-classic-benchmark
+        files: ^classic/benchmark/(agbenchmark|tests)/((?!reports).)*[/.]
+        args: [--config=classic/benchmark/.flake8]
 
   - repo: local
     hooks:
@@ -253,10 +311,29 @@ repos:
         pass_filenames: false
 
       - id: pyright
-        name: Typecheck - Classic
-        alias: pyright-classic
-        entry: poetry -C classic run pyright
-        files: ^classic/(original_autogpt|forge|direct_benchmark)/.*\.py$|^classic/poetry\.lock$
+        name: Typecheck - Classic - AutoGPT
+        alias: pyright-classic-autogpt
+        entry: poetry -C classic/original_autogpt run pyright
+        # include forge source (since it's a path dependency) but exclude *_test.py files:
+        files: ^(classic/original_autogpt/((autogpt|scripts|tests)/|poetry\.lock$)|classic/forge/(forge/.*(?<!_test)\.py|poetry\.lock)$)
+        types: [file]
+        language: system
+        pass_filenames: false
+
+      - id: pyright
+        name: Typecheck - Classic - Forge
+        alias: pyright-classic-forge
+        entry: poetry -C classic/forge run pyright
+        files: ^classic/forge/(forge/|poetry\.lock$)
+        types: [file]
+        language: system
+        pass_filenames: false
+
+      - id: pyright
+        name: Typecheck - Classic - Benchmark
+        alias: pyright-classic-benchmark
+        entry: poetry -C classic/benchmark run pyright
+        files: ^classic/benchmark/(agbenchmark/|tests/|poetry\.lock$)
         types: [file]
         language: system
         pass_filenames: false
@@ -283,9 +360,26 @@ repos:
   #       pass_filenames: false
 
   #     - id: pytest
-  #       name: Run tests - Classic (excl. slow tests)
-  #       alias: pytest-classic
-  #       entry: bash -c 'cd classic && poetry run pytest -m "not slow"'
-  #       files: ^classic/(original_autogpt|forge|direct_benchmark)/
+  #       name: Run tests - Classic - AutoGPT (excl. slow tests)
+  #       alias: pytest-classic-autogpt
+  #       entry: bash -c 'cd classic/original_autogpt && poetry run pytest --cov=autogpt -m "not slow" tests/unit tests/integration'
+  #       # include forge source (since it's a path dependency) but exclude *_test.py files:
+  #       files: ^(classic/original_autogpt/((autogpt|tests)/|poetry\.lock$)|classic/forge/(forge/.*(?<!_test)\.py|poetry\.lock)$)
+  #       language: system
+  #       pass_filenames: false
+
+  #     - id: pytest
+  #       name: Run tests - Classic - Forge (excl. slow tests)
+  #       alias: pytest-classic-forge
+  #       entry: bash -c 'cd classic/forge && poetry run pytest --cov=forge -m "not slow"'
+  #       files: ^classic/forge/(forge/|tests/|poetry\.lock$)
+  #       language: system
+  #       pass_filenames: false
+
+  #     - id: pytest
+  #       name: Run tests - Classic - Benchmark
+  #       alias: pytest-classic-benchmark
+  #       entry: bash -c 'cd classic/benchmark && poetry run pytest --cov=benchmark'
+  #       files: ^classic/benchmark/(agbenchmark/|tests/|poetry\.lock$)
   #       language: system
   #       pass_filenames: false

.secrets.baseline

@@ -1,467 +0,0 @@
-{
-  "version": "1.5.0",
-  "plugins_used": [
-    {
-      "name": "ArtifactoryDetector"
-    },
-    {
-      "name": "AWSKeyDetector"
-    },
-    {
-      "name": "AzureStorageKeyDetector"
-    },
-    {
-      "name": "Base64HighEntropyString",
-      "limit": 4.5
-    },
-    {
-      "name": "BasicAuthDetector"
-    },
-    {
-      "name": "CloudantDetector"
-    },
-    {
-      "name": "DiscordBotTokenDetector"
-    },
-    {
-      "name": "GitHubTokenDetector"
-    },
-    {
-      "name": "GitLabTokenDetector"
-    },
-    {
-      "name": "HexHighEntropyString",
-      "limit": 3.0
-    },
-    {
-      "name": "IbmCloudIamDetector"
-    },
-    {
-      "name": "IbmCosHmacDetector"
-    },
-    {
-      "name": "IPPublicDetector"
-    },
-    {
-      "name": "JwtTokenDetector"
-    },
-    {
-      "name": "KeywordDetector",
-      "keyword_exclude": ""
-    },
-    {
-      "name": "MailchimpDetector"
-    },
-    {
-      "name": "NpmDetector"
-    },
-    {
-      "name": "OpenAIDetector"
-    },
-    {
-      "name": "PrivateKeyDetector"
-    },
-    {
-      "name": "PypiTokenDetector"
-    },
-    {
-      "name": "SendGridDetector"
-    },
-    {
-      "name": "SlackDetector"
-    },
-    {
-      "name": "SoftlayerDetector"
-    },
-    {
-      "name": "SquareOAuthDetector"
-    },
-    {
-      "name": "StripeDetector"
-    },
-    {
-      "name": "TelegramBotTokenDetector"
-    },
-    {
-      "name": "TwilioKeyDetector"
-    }
-  ],
-  "filters_used": [
-    {
-      "path": "detect_secrets.filters.allowlist.is_line_allowlisted"
-    },
-    {
-      "path": "detect_secrets.filters.common.is_ignored_due_to_verification_policies",
-      "min_level": 2
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_indirect_reference"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_likely_id_string"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_lock_file"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_not_alphanumeric_string"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_potential_uuid"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_prefixed_with_dollar_sign"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_sequential_string"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_swagger_file"
-    },
-    {
-      "path": "detect_secrets.filters.heuristic.is_templated_secret"
-    },
-    {
-      "path": "detect_secrets.filters.regex.should_exclude_file",
-      "pattern": [
-        "\\.env$",
-        "pnpm-lock\\.yaml$",
-        "\\.env\\.(default|example|template)$",
-        "__pycache__",
-        "_test\\.py$",
-        "test_.*\\.py$",
-        "conftest\\.py$",
-        "poetry\\.lock$",
-        "node_modules"
-      ]
-    }
-  ],
-  "results": {
-    "autogpt_platform/backend/backend/api/external/v1/integrations.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/api/external/v1/integrations.py",
-        "hashed_secret": "665b1e3851eefefa3fb878654292f16597d25155",
-        "is_verified": false,
-        "line_number": 289
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/airtable/_config.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/blocks/airtable/_config.py",
-        "hashed_secret": "57e168b03afb7c1ee3cdc4ee3db2fe1cc6e0df26",
-        "is_verified": false,
-        "line_number": 29
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/dataforseo/_config.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/blocks/dataforseo/_config.py",
-        "hashed_secret": "32ce93887331fa5d192f2876ea15ec000c7d58b8",
-        "is_verified": false,
-        "line_number": 12
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/github/checks.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/checks.py",
-        "hashed_secret": "8ac6f92737d8586790519c5d7bfb4d2eb172c238",
-        "is_verified": false,
-        "line_number": 108
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/github/ci.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/ci.py",
-        "hashed_secret": "90bd1b48e958257948487b90bee080ba5ed00caa",
-        "is_verified": false,
-        "line_number": 123
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/github/example_payloads/pull_request.synchronize.json": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/example_payloads/pull_request.synchronize.json",
-        "hashed_secret": "f96896dafced7387dcd22343b8ea29d3d2c65663",
-        "is_verified": false,
-        "line_number": 42
-      },
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/example_payloads/pull_request.synchronize.json",
-        "hashed_secret": "b80a94d5e70bedf4f5f89d2f5a5255cc9492d12e",
-        "is_verified": false,
-        "line_number": 193
-      },
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/example_payloads/pull_request.synchronize.json",
-        "hashed_secret": "75b17e517fe1b3136394f6bec80c4f892da75e42",
-        "is_verified": false,
-        "line_number": 344
-      },
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/example_payloads/pull_request.synchronize.json",
-        "hashed_secret": "b0bfb5e4e2394e7f8906e5ed1dffd88b2bc89dd5",
-        "is_verified": false,
-        "line_number": 534
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/github/statuses.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/github/statuses.py",
-        "hashed_secret": "8ac6f92737d8586790519c5d7bfb4d2eb172c238",
-        "is_verified": false,
-        "line_number": 85
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/google/docs.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/google/docs.py",
-        "hashed_secret": "c95da0c6696342c867ef0c8258d2f74d20fd94d4",
-        "is_verified": false,
-        "line_number": 203
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/google/sheets.py": [
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/google/sheets.py",
-        "hashed_secret": "bd5a04fa3667e693edc13239b6d310c5c7a8564b",
-        "is_verified": false,
-        "line_number": 57
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/linear/_config.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/blocks/linear/_config.py",
-        "hashed_secret": "b37f020f42d6d613b6ce30103e4d408c4499b3bb",
-        "is_verified": false,
-        "line_number": 53
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/medium.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/medium.py",
-        "hashed_secret": "ff998abc1ce6d8f01a675fa197368e44c8916e9c",
-        "is_verified": false,
-        "line_number": 131
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/replicate/replicate_block.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/replicate/replicate_block.py",
-        "hashed_secret": "8bbdd6f26368f58ea4011d13d7f763cb662e66f0",
-        "is_verified": false,
-        "line_number": 55
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/slant3d/webhook.py": [
-      {
-        "type": "Hex High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/slant3d/webhook.py",
-        "hashed_secret": "36263c76947443b2f6e6b78153967ac4a7da99f9",
-        "is_verified": false,
-        "line_number": 100
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/talking_head.py": [
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/backend/backend/blocks/talking_head.py",
-        "hashed_secret": "44ce2d66222529eea4a32932823466fc0601c799",
-        "is_verified": false,
-        "line_number": 113
-      }
-    ],
-    "autogpt_platform/backend/backend/blocks/wordpress/_config.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/blocks/wordpress/_config.py",
-        "hashed_secret": "e62679512436161b78e8a8d68c8829c2a1031ccb",
-        "is_verified": false,
-        "line_number": 17
-      }
-    ],
-    "autogpt_platform/backend/backend/util/cache.py": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/backend/backend/util/cache.py",
-        "hashed_secret": "37f0c918c3fa47ca4a70e42037f9f123fdfbc75b",
-        "is_verified": false,
-        "line_number": 449
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/build/components/FlowEditor/nodes/helpers.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/build/components/FlowEditor/nodes/helpers.ts",
-        "hashed_secret": "5baa61e4c9b93f3f0682250b6cf8331b7ee68fd8",
-        "is_verified": false,
-        "line_number": 6
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/dictionaries/en.json": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/dictionaries/en.json",
-        "hashed_secret": "8be3c943b1609fffbfc51aad666d0a04adf83c9d",
-        "is_verified": false,
-        "line_number": 5
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/dictionaries/es.json": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/dictionaries/es.json",
-        "hashed_secret": "5a6d1c612954979ea99ee33dbb2d231b00f6ac0a",
-        "is_verified": false,
-        "line_number": 5
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/AgentInputsReadOnly/helpers.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/AgentInputsReadOnly/helpers.ts",
-        "hashed_secret": "cf678cab87dc1f7d1b95b964f15375e088461679",
-        "is_verified": false,
-        "line_number": 6
-      },
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/AgentInputsReadOnly/helpers.ts",
-        "hashed_secret": "f72cbb45464d487064610c5411c576ca4019d380",
-        "is_verified": false,
-        "line_number": 8
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/RunAgentModal/components/ModalRunSection/helpers.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/RunAgentModal/components/ModalRunSection/helpers.ts",
-        "hashed_secret": "cf678cab87dc1f7d1b95b964f15375e088461679",
-        "is_verified": false,
-        "line_number": 5
-      },
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/library/agents/[id]/components/NewAgentLibraryView/components/modals/RunAgentModal/components/ModalRunSection/helpers.ts",
-        "hashed_secret": "f72cbb45464d487064610c5411c576ca4019d380",
-        "is_verified": false,
-        "line_number": 7
-      }
-    ],
-    "autogpt_platform/frontend/src/app/(platform)/profile/(user)/integrations/page.tsx": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/profile/(user)/integrations/page.tsx",
-        "hashed_secret": "cf678cab87dc1f7d1b95b964f15375e088461679",
-        "is_verified": false,
-        "line_number": 192
-      },
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/app/(platform)/profile/(user)/integrations/page.tsx",
-        "hashed_secret": "86275db852204937bbdbdebe5fabe8536e030ab6",
-        "is_verified": false,
-        "line_number": 193
-      }
-    ],
-    "autogpt_platform/frontend/src/components/contextual/CredentialsInput/helpers.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/components/contextual/CredentialsInput/helpers.ts",
-        "hashed_secret": "47acd2028cf81b5da88ddeedb2aea4eca4b71fbd",
-        "is_verified": false,
-        "line_number": 102
-      },
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/components/contextual/CredentialsInput/helpers.ts",
-        "hashed_secret": "8be3c943b1609fffbfc51aad666d0a04adf83c9d",
-        "is_verified": false,
-        "line_number": 103
-      }
-    ],
-    "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts": [
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "9c486c92f1a7420e1045c7ad963fbb7ba3621025",
-        "is_verified": false,
-        "line_number": 73
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "9277508c7a6effc8fb59163efbfada189e35425c",
-        "is_verified": false,
-        "line_number": 75
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "8dc7e2cb1d0935897d541bf5facab389b8a50340",
-        "is_verified": false,
-        "line_number": 77
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "79a26ad48775944299be6aaf9fb1d5302c1ed75b",
-        "is_verified": false,
-        "line_number": 79
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "a3b62b44500a1612e48d4cab8294df81561b3b1a",
-        "is_verified": false,
-        "line_number": 81
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "a58979bd0b21ef4f50417d001008e60dd7a85c64",
-        "is_verified": false,
-        "line_number": 83
-      },
-      {
-        "type": "Base64 High Entropy String",
-        "filename": "autogpt_platform/frontend/src/lib/autogpt-server-api/utils.ts",
-        "hashed_secret": "6cb6e075f8e8c7c850f9d128d6608e5dbe209a79",
-        "is_verified": false,
-        "line_number": 85
-      }
-    ],
-    "autogpt_platform/frontend/src/lib/constants.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/lib/constants.ts",
-        "hashed_secret": "27b924db06a28cc755fb07c54f0fddc30659fe4d",
-        "is_verified": false,
-        "line_number": 10
-      }
-    ],
-    "autogpt_platform/frontend/src/tests/credentials/index.ts": [
-      {
-        "type": "Secret Keyword",
-        "filename": "autogpt_platform/frontend/src/tests/credentials/index.ts",
-        "hashed_secret": "c18006fc138809314751cd1991f1e0b820fabd37",
-        "is_verified": false,
-        "line_number": 4
-      }
-    ]
-  },
-  "generated_at": "2026-04-02T13:10:54Z"
-}

AGENTS.md

@@ -30,7 +30,7 @@ See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference:
    - Regenerate with `pnpm generate:api`
    - Pattern: `use{Method}{Version}{OperationName}`
 4. **Styling**: Tailwind CSS only, use design tokens, Phosphor Icons only
-5. **Testing**: Integration tests (Vitest + RTL + MSW) are the default (~90%, page-level). Playwright for E2E critical flows. Storybook for design system components. See `autogpt_platform/frontend/TESTING.md`
+5. **Testing**: Add Storybook stories for new components, Playwright for E2E
 6. **Code conventions**: Function declarations (not arrow functions) for components/handlers
 
 - Component props should be `interface Props { ... }` (not exported) unless the interface needs to be used outside the component
@@ -47,9 +47,7 @@ See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference:
 ## Testing
 
 - Backend: `poetry run test` (runs pytest with a docker based postgres + prisma).
-- Frontend integration tests: `pnpm test:unit` (Vitest + RTL + MSW, primary testing approach).
-- Frontend E2E tests: `pnpm test` or `pnpm test-ui` for Playwright tests.
-- See `autogpt_platform/frontend/TESTING.md` for the full testing strategy.
+- Frontend: `pnpm test` or `pnpm test-ui` for Playwright tests. See `docs/content/platform/contributing/tests.md` for tips.
 
 Always run the relevant linters and tests before committing.
 Use conventional commit messages for all commits (e.g. `feat(backend): add API`).

autogpt_platform/backend/backend/api/features/admin/platform_cost_routes.py

@@ -0,0 +1,85 @@
+import logging
+import typing
+from datetime import datetime
+
+from autogpt_libs.auth import get_user_id, requires_admin_user
+from fastapi import APIRouter, Query, Security
+from pydantic import BaseModel
+
+from backend.data.platform_cost import (
+    CostLogRow,
+    PlatformCostDashboard,
+    get_platform_cost_dashboard,
+    get_platform_cost_logs,
+)
+from backend.util.models import Pagination
+
+logger = logging.getLogger(__name__)
+
+
+router = APIRouter(
+    prefix="/admin",
+    tags=["platform-cost", "admin"],
+    dependencies=[Security(requires_admin_user)],
+)
+
+
+class PlatformCostLogsResponse(BaseModel):
+    logs: list[CostLogRow]
+    pagination: Pagination
+
+
+@router.get(
+    "/platform_costs/dashboard",
+    response_model=PlatformCostDashboard,
+    summary="Get Platform Cost Dashboard",
+)
+async def get_cost_dashboard(
+    admin_user_id: str = Security(get_user_id),
+    start: typing.Optional[datetime] = Query(None),
+    end: typing.Optional[datetime] = Query(None),
+    provider: typing.Optional[str] = Query(None),
+    user_id: typing.Optional[str] = Query(None),
+):
+    logger.info(f"Admin {admin_user_id} fetching platform cost dashboard")
+    return await get_platform_cost_dashboard(
+        start=start,
+        end=end,
+        provider=provider,
+        user_id=user_id,
+    )
+
+
+@router.get(
+    "/platform_costs/logs",
+    response_model=PlatformCostLogsResponse,
+    summary="Get Platform Cost Logs",
+)
+async def get_cost_logs(
+    admin_user_id: str = Security(get_user_id),
+    start: typing.Optional[datetime] = Query(None),
+    end: typing.Optional[datetime] = Query(None),
+    provider: typing.Optional[str] = Query(None),
+    user_id: typing.Optional[str] = Query(None),
+    page: int = Query(1, ge=1),
+    page_size: int = Query(50, ge=1, le=200),
+):
+    logger.info(f"Admin {admin_user_id} fetching platform cost logs")
+    logs, total = await get_platform_cost_logs(
+        start=start,
+        end=end,
+        provider=provider,
+        user_id=user_id,
+        page=page,
+        page_size=page_size,
+    )
+    total_pages = (total + page_size - 1) // page_size
+    return PlatformCostLogsResponse(
+        logs=logs,
+        pagination=Pagination(
+            total_items=total,
+            total_pages=total_pages,
+            current_page=page,
+            page_size=page_size,
+        ),
+    )

autogpt_platform/backend/backend/api/features/admin/platform_cost_routes_test.py

@@ -0,0 +1,135 @@
+from unittest.mock import AsyncMock
+
+import fastapi
+import fastapi.testclient
+import pytest
+import pytest_mock
+from autogpt_libs.auth.jwt_utils import get_jwt_payload
+
+from .platform_cost_routes import router as platform_cost_router
+
+app = fastapi.FastAPI()
+app.include_router(platform_cost_router)
+
+client = fastapi.testclient.TestClient(app)
+
+
+@pytest.fixture(autouse=True)
+def setup_app_admin_auth(mock_jwt_admin):
+    """Setup admin auth overrides for all tests in this module"""
+    app.dependency_overrides[get_jwt_payload] = mock_jwt_admin["get_jwt_payload"]
+    yield
+    app.dependency_overrides.clear()
+
+
+def test_get_dashboard_success(
+    mocker: pytest_mock.MockerFixture,
+) -> None:
+    mock_dashboard = AsyncMock(
+        return_value=AsyncMock(
+            by_provider=[],
+            by_user=[],
+            total_cost_microdollars=0,
+            total_requests=0,
+            total_users=0,
+            model_dump=lambda **_: {
+                "by_provider": [],
+                "by_user": [],
+                "total_cost_microdollars": 0,
+                "total_requests": 0,
+                "total_users": 0,
+            },
+        )
+    )
+    mocker.patch(
+        "backend.api.features.admin.platform_cost_routes.get_platform_cost_dashboard",
+        mock_dashboard,
+    )
+
+    response = client.get("/admin/platform_costs/dashboard")
+    assert response.status_code == 200
+    data = response.json()
+    assert "by_provider" in data
+    assert "by_user" in data
+    assert data["total_cost_microdollars"] == 0
+
+
+def test_get_logs_success(
+    mocker: pytest_mock.MockerFixture,
+) -> None:
+    mocker.patch(
+        "backend.api.features.admin.platform_cost_routes.get_platform_cost_logs",
+        AsyncMock(return_value=([], 0)),
+    )
+
+    response = client.get("/admin/platform_costs/logs")
+    assert response.status_code == 200
+    data = response.json()
+    assert data["logs"] == []
+    assert data["pagination"]["total_items"] == 0
+
+
+def test_get_dashboard_with_filters(
+    mocker: pytest_mock.MockerFixture,
+) -> None:
+    mock_dashboard = AsyncMock(
+        return_value=AsyncMock(
+            by_provider=[],
+            by_user=[],
+            total_cost_microdollars=0,
+            total_requests=0,
+            total_users=0,
+            model_dump=lambda **_: {
+                "by_provider": [],
+                "by_user": [],
+                "total_cost_microdollars": 0,
+                "total_requests": 0,
+                "total_users": 0,
+            },
+        )
+    )
+    mocker.patch(
+        "backend.api.features.admin.platform_cost_routes.get_platform_cost_dashboard",
+        mock_dashboard,
+    )
+
+    response = client.get(
+        "/admin/platform_costs/dashboard",
+        params={
+            "start": "2026-01-01T00:00:00",
+            "end": "2026-04-01T00:00:00",
+            "provider": "openai",
+            "user_id": "test-user-123",
+        },
+    )
+    assert response.status_code == 200
+    mock_dashboard.assert_called_once()
+    call_kwargs = mock_dashboard.call_args.kwargs
+    assert call_kwargs["provider"] == "openai"
+    assert call_kwargs["user_id"] == "test-user-123"
+    assert call_kwargs["start"] is not None
+    assert call_kwargs["end"] is not None
+
+
+def test_get_logs_with_pagination(
+    mocker: pytest_mock.MockerFixture,
+) -> None:
+    mocker.patch(
+        "backend.api.features.admin.platform_cost_routes.get_platform_cost_logs",
+        AsyncMock(return_value=([], 0)),
+    )
+
+    response = client.get(
+        "/admin/platform_costs/logs",
+        params={"page": 2, "page_size": 25, "provider": "anthropic"},
+    )
+    assert response.status_code == 200
+    data = response.json()
+    assert data["pagination"]["current_page"] == 2
+    assert data["pagination"]["page_size"] == 25
+
+
+def test_get_dashboard_requires_admin() -> None:
+    app.dependency_overrides.clear()
+    response = client.get("/admin/platform_costs/dashboard")
+    assert response.status_code in (401, 403)

autogpt_platform/backend/backend/api/features/admin/rate_limit_admin_routes.py

@@ -194,29 +194,23 @@ async def set_user_rate_limit_tier(
     request: SetUserTierRequest,
     admin_user_id: str = Security(get_user_id),
 ) -> UserTierResponse:
-    """Set a user's rate-limit tier. Admin-only.
+    """Set a user's rate-limit tier. Admin-only."""
+    old_tier = await get_user_tier(request.user_id)
 
-    Returns 404 if the user does not exist in the database.
-    """
+    # Resolve email for audit logging (non-blocking — don't fail the
+    # tier change if email lookup fails).
     try:
         resolved_email = await get_user_email_by_id(request.user_id)
     except Exception:
         logger.warning(
-            "Failed to resolve email for user %s",
-            request.user_id,
-            exc_info=True,
+            "Failed to resolve email for user %s", request.user_id, exc_info=True
         )
         resolved_email = None
-
-    if resolved_email is None:
-        raise HTTPException(status_code=404, detail=f"User {request.user_id} not found")
-
-    old_tier = await get_user_tier(request.user_id)
     logger.info(
         "Admin %s changing tier for user %s (%s): %s -> %s",
         admin_user_id,
         request.user_id,
-        resolved_email,
+        resolved_email or "unknown",
         old_tier.value,
         request.tier.value,
     )

autogpt_platform/backend/backend/api/features/admin/rate_limit_admin_routes_test.py

@@ -405,34 +405,24 @@ def test_set_user_tier_invalid_tier_uppercase(
     assert "detail" in body
 
 
-def test_set_user_tier_email_lookup_failure_returns_404(
+def test_set_user_tier_email_lookup_failure_non_blocking(
     mocker: pytest_mock.MockerFixture,
     target_user_id: str,
 ) -> None:
-    """Test that email lookup failure returns 404 (user unverifiable)."""
+    """Test that email lookup failure doesn't block tier change."""
     mocker.patch(
         f"{_MOCK_MODULE}.get_user_email_by_id",
         new_callable=AsyncMock,
         side_effect=Exception("DB connection failed"),
     )
-
-    response = client.post(
-        "/admin/rate_limit/tier",
-        json={"user_id": target_user_id, "tier": "PRO"},
-    )
-
-    assert response.status_code == 404
-
-
-def test_set_user_tier_user_not_found(
-    mocker: pytest_mock.MockerFixture,
-    target_user_id: str,
-) -> None:
-    """Test that setting tier for a non-existent user returns 404."""
     mocker.patch(
-        f"{_MOCK_MODULE}.get_user_email_by_id",
+        f"{_MOCK_MODULE}.get_user_tier",
+        new_callable=AsyncMock,
+        return_value=SubscriptionTier.FREE,
+    )
+    mock_set = mocker.patch(
+        f"{_MOCK_MODULE}.set_user_tier",
         new_callable=AsyncMock,
-        return_value=None,
     )
 
     response = client.post(
@@ -440,7 +430,8 @@ def test_set_user_tier_user_not_found(
         json={"user_id": target_user_id, "tier": "PRO"},
     )
 
-    assert response.status_code == 404
+    assert response.status_code == 200
+    mock_set.assert_awaited_once()
 
 
 def test_set_user_tier_db_failure(

autogpt_platform/backend/backend/api/features/chat/routes.py

@@ -4,7 +4,7 @@ import asyncio
 import logging
 import re
 from collections.abc import AsyncGenerator
-from typing import Annotated
+from typing import Annotated, Literal
 from uuid import uuid4
 
 from autogpt_libs import auth
@@ -15,8 +15,7 @@ from pydantic import BaseModel, ConfigDict, Field, field_validator
 
 from backend.copilot import service as chat_service
 from backend.copilot import stream_registry
-from backend.copilot.config import ChatConfig, CopilotMode
-from backend.copilot.db import get_chat_messages_paginated
+from backend.copilot.config import ChatConfig
 from backend.copilot.executor.utils import enqueue_cancel_task, enqueue_copilot_turn
 from backend.copilot.model import (
     ChatMessage,
@@ -112,7 +111,7 @@ class StreamChatRequest(BaseModel):
     file_ids: list[str] | None = Field(
         default=None, max_length=20
     )  # Workspace file IDs attached to this message
-    mode: CopilotMode | None = Field(
+    mode: Literal["fast", "extended_thinking"] | None = Field(
         default=None,
         description="Autopilot mode: 'fast' for baseline LLM, 'extended_thinking' for Claude Agent SDK. "
         "If None, uses the server default (extended_thinking).",
@@ -156,8 +155,6 @@ class SessionDetailResponse(BaseModel):
     user_id: str | None
     messages: list[dict]
     active_stream: ActiveStreamInfo | None = None  # Present if stream is still active
-    has_more_messages: bool = False
-    oldest_sequence: int | None = None
     total_prompt_tokens: int = 0
     total_completion_tokens: int = 0
     metadata: ChatSessionMetadata = ChatSessionMetadata()
@@ -397,78 +394,60 @@ async def update_session_title_route(
 async def get_session(
     session_id: str,
     user_id: Annotated[str, Security(auth.get_user_id)],
-    limit: int = Query(default=50, ge=1, le=200),
-    before_sequence: int | None = Query(default=None, ge=0),
 ) -> SessionDetailResponse:
     """
     Retrieve the details of a specific chat session.
 
-    Supports cursor-based pagination via ``limit`` and ``before_sequence``.
-    When no pagination params are provided, returns the most recent messages.
+    Looks up a chat session by ID for the given user (if authenticated) and returns all session data including messages.
+    If there's an active stream for this session, returns active_stream info for reconnection.
 
     Args:
         session_id: The unique identifier for the desired chat session.
-        user_id: The authenticated user's ID.
-        limit: Maximum number of messages to return (1-200, default 50).
-        before_sequence: Return messages with sequence < this value (cursor).
+        user_id: The optional authenticated user ID, or None for anonymous access.
 
     Returns:
-        SessionDetailResponse: Details for the requested session, including
-            active_stream info and pagination metadata.
+        SessionDetailResponse: Details for the requested session, including active_stream info if applicable.
+
     """
-    page = await get_chat_messages_paginated(
-        session_id, limit, before_sequence, user_id=user_id
-    )
-    if page is None:
+    session = await get_chat_session(session_id, user_id)
+    if not session:
         raise NotFoundError(f"Session {session_id} not found.")
-    messages = [message.model_dump() for message in page.messages]
 
-    # Only check active stream on initial load (not on "load more" requests)
+    messages = [message.model_dump() for message in session.messages]
+
+    # Check if there's an active stream for this session
     active_stream_info = None
-    if before_sequence is None:
-        active_session, last_message_id = await stream_registry.get_active_session(
-            session_id, user_id
-        )
-        logger.info(
-            f"[GET_SESSION] session={session_id}, active_session={active_session is not None}, "
-            f"msg_count={len(messages)}, last_role={messages[-1].get('role') if messages else 'none'}"
-        )
-        if active_session:
-            active_stream_info = ActiveStreamInfo(
-                turn_id=active_session.turn_id,
-                last_message_id=last_message_id,
-            )
-
-    # Skip session metadata on "load more" — frontend only needs messages
-    if before_sequence is not None:
-        return SessionDetailResponse(
-            id=page.session.session_id,
-            created_at=page.session.started_at.isoformat(),
-            updated_at=page.session.updated_at.isoformat(),
-            user_id=page.session.user_id or None,
-            messages=messages,
-            active_stream=None,
-            has_more_messages=page.has_more,
-            oldest_sequence=page.oldest_sequence,
-            total_prompt_tokens=0,
-            total_completion_tokens=0,
+    active_session, last_message_id = await stream_registry.get_active_session(
+        session_id, user_id
+    )
+    logger.info(
+        f"[GET_SESSION] session={session_id}, active_session={active_session is not None}, "
+        f"msg_count={len(messages)}, last_role={messages[-1].get('role') if messages else 'none'}"
+    )
+    if active_session:
+        # Keep the assistant message (including tool_calls) so the frontend can
+        # render the correct tool UI (e.g. CreateAgent with mini game).
+        # convertChatSessionToUiMessages handles isComplete=false by setting
+        # tool parts without output to state "input-available".
+        active_stream_info = ActiveStreamInfo(
+            turn_id=active_session.turn_id,
+            last_message_id=last_message_id,
         )
 
-    total_prompt = sum(u.prompt_tokens for u in page.session.usage)
-    total_completion = sum(u.completion_tokens for u in page.session.usage)
+    # Sum token usage from session
+    total_prompt = sum(u.prompt_tokens for u in session.usage)
+    total_completion = sum(u.completion_tokens for u in session.usage)
 
     return SessionDetailResponse(
-        id=page.session.session_id,
-        created_at=page.session.started_at.isoformat(),
-        updated_at=page.session.updated_at.isoformat(),
-        user_id=page.session.user_id or None,
+        id=session.session_id,
+        created_at=session.started_at.isoformat(),
+        updated_at=session.updated_at.isoformat(),
+        user_id=session.user_id or None,
         messages=messages,
         active_stream=active_stream_info,
-        has_more_messages=page.has_more,
-        oldest_sequence=page.oldest_sequence,
         total_prompt_tokens=total_prompt,
         total_completion_tokens=total_completion,
-        metadata=page.session.metadata,
+        metadata=session.metadata,
     )
 
 

autogpt_platform/backend/backend/api/features/chat/routes_test.py

@@ -541,41 +541,3 @@ def test_create_session_rejects_nested_metadata(
     )
 
     assert response.status_code == 422
-
-
-class TestStreamChatRequestModeValidation:
-    """Pydantic-level validation of the ``mode`` field on StreamChatRequest."""
-
-    def test_rejects_invalid_mode_value(self) -> None:
-        """Any string outside the Literal set must raise ValidationError."""
-        from pydantic import ValidationError
-
-        from backend.api.features.chat.routes import StreamChatRequest
-
-        with pytest.raises(ValidationError):
-            StreamChatRequest(message="hi", mode="turbo")  # type: ignore[arg-type]
-
-    def test_accepts_fast_mode(self) -> None:
-        from backend.api.features.chat.routes import StreamChatRequest
-
-        req = StreamChatRequest(message="hi", mode="fast")
-        assert req.mode == "fast"
-
-    def test_accepts_extended_thinking_mode(self) -> None:
-        from backend.api.features.chat.routes import StreamChatRequest
-
-        req = StreamChatRequest(message="hi", mode="extended_thinking")
-        assert req.mode == "extended_thinking"
-
-    def test_accepts_none_mode(self) -> None:
-        """``mode=None`` is valid (server decides via feature flags)."""
-        from backend.api.features.chat.routes import StreamChatRequest
-
-        req = StreamChatRequest(message="hi", mode=None)
-        assert req.mode is None
-
-    def test_mode_defaults_to_none_when_omitted(self) -> None:
-        from backend.api.features.chat.routes import StreamChatRequest
-
-        req = StreamChatRequest(message="hi")
-        assert req.mode is None

autogpt_platform/backend/backend/api/features/onboarding_profile_test.py

@@ -1,61 +0,0 @@
-from unittest.mock import AsyncMock
-
-import fastapi
-import fastapi.testclient
-import pytest
-
-from backend.api.features.v1 import v1_router
-
-app = fastapi.FastAPI()
-app.include_router(v1_router)
-client = fastapi.testclient.TestClient(app)
-
-
-@pytest.fixture(autouse=True)
-def setup_app_auth(mock_jwt_user):
-    from autogpt_libs.auth.jwt_utils import get_jwt_payload
-
-    app.dependency_overrides[get_jwt_payload] = mock_jwt_user["get_jwt_payload"]
-    yield
-    app.dependency_overrides.clear()
-
-
-def test_onboarding_profile_success(mocker):
-    mock_extract = mocker.patch(
-        "backend.api.features.v1.extract_business_understanding",
-        new_callable=AsyncMock,
-    )
-    mock_upsert = mocker.patch(
-        "backend.api.features.v1.upsert_business_understanding",
-        new_callable=AsyncMock,
-    )
-
-    from backend.data.understanding import BusinessUnderstandingInput
-
-    mock_extract.return_value = BusinessUnderstandingInput.model_construct(
-        user_name="John",
-        user_role="Founder/CEO",
-        pain_points=["Finding leads"],
-        suggested_prompts={"Learn": ["How do I automate lead gen?"]},
-    )
-    mock_upsert.return_value = AsyncMock()
-
-    response = client.post(
-        "/onboarding/profile",
-        json={
-            "user_name": "John",
-            "user_role": "Founder/CEO",
-            "pain_points": ["Finding leads", "Email & outreach"],
-        },
-    )
-    assert response.status_code == 200
-    mock_extract.assert_awaited_once()
-    mock_upsert.assert_awaited_once()
-
-
-def test_onboarding_profile_missing_fields():
-    response = client.post(
-        "/onboarding/profile",
-        json={"user_name": "John"},
-    )
-    assert response.status_code == 422

autogpt_platform/backend/backend/api/features/v1.py

@@ -63,17 +63,12 @@ from backend.data.onboarding import (
     UserOnboardingUpdate,
     complete_onboarding_step,
     complete_re_run_agent,
-    format_onboarding_for_extraction,
     get_recommended_agents,
     get_user_onboarding,
+    onboarding_enabled,
     reset_user_onboarding,
     update_user_onboarding,
 )
-from backend.data.tally import extract_business_understanding
-from backend.data.understanding import (
-    BusinessUnderstandingInput,
-    upsert_business_understanding,
-)
 from backend.data.user import (
     get_or_create_user,
     get_user_by_id,
@@ -287,33 +282,35 @@ async def get_onboarding_agents(
     return await get_recommended_agents(user_id)
 
 
-class OnboardingProfileRequest(pydantic.BaseModel):
-    """Request body for onboarding profile submission."""
-
-    user_name: str = pydantic.Field(min_length=1, max_length=100)
-    user_role: str = pydantic.Field(min_length=1, max_length=100)
-    pain_points: list[str] = pydantic.Field(default_factory=list, max_length=20)
-
-
 class OnboardingStatusResponse(pydantic.BaseModel):
-    """Response for onboarding completion check."""
+    """Response for onboarding status check."""
 
-    is_completed: bool
+    is_onboarding_enabled: bool
+    is_chat_enabled: bool
 
 
 @v1_router.get(
-    "/onboarding/completed",
-    summary="Check if onboarding is completed",
+    "/onboarding/enabled",
+    summary="Is onboarding enabled",
     tags=["onboarding", "public"],
     response_model=OnboardingStatusResponse,
-    dependencies=[Security(requires_user)],
 )
-async def is_onboarding_completed(
+async def is_onboarding_enabled(
     user_id: Annotated[str, Security(get_user_id)],
 ) -> OnboardingStatusResponse:
-    user_onboarding = await get_user_onboarding(user_id)
+    # Check if chat is enabled for user
+    is_chat_enabled = await is_feature_enabled(Flag.CHAT, user_id, False)
+
+    # If chat is enabled, skip legacy onboarding
+    if is_chat_enabled:
+        return OnboardingStatusResponse(
+            is_onboarding_enabled=False,
+            is_chat_enabled=True,
+        )
+
     return OnboardingStatusResponse(
-        is_completed=OnboardingStep.VISIT_COPILOT in user_onboarding.completedSteps,
+        is_onboarding_enabled=await onboarding_enabled(),
+        is_chat_enabled=False,
     )
 
 
@@ -328,38 +325,6 @@ async def reset_onboarding(user_id: Annotated[str, Security(get_user_id)]):
     return await reset_user_onboarding(user_id)
 
 
-@v1_router.post(
-    "/onboarding/profile",
-    summary="Submit onboarding profile",
-    tags=["onboarding"],
-    dependencies=[Security(requires_user)],
-)
-async def submit_onboarding_profile(
-    data: OnboardingProfileRequest,
-    user_id: Annotated[str, Security(get_user_id)],
-):
-    formatted = format_onboarding_for_extraction(
-        user_name=data.user_name,
-        user_role=data.user_role,
-        pain_points=data.pain_points,
-    )
-
-    try:
-        understanding_input = await extract_business_understanding(formatted)
-    except Exception:
-        understanding_input = BusinessUnderstandingInput.model_construct()
-
-    # Ensure the direct fields are set even if LLM missed them
-    understanding_input.user_name = data.user_name
-    understanding_input.user_role = data.user_role
-    if not understanding_input.pain_points:
-        understanding_input.pain_points = data.pain_points
-
-    await upsert_business_understanding(user_id, understanding_input)
-
-    return {"status": "ok"}
-
-
 ########################################################
 ##################### Blocks ###########################
 ########################################################

autogpt_platform/backend/backend/api/features/workspace/routes.py

@@ -12,7 +12,7 @@ import fastapi
 from autogpt_libs.auth.dependencies import get_user_id, requires_user
 from fastapi import Query, UploadFile
 from fastapi.responses import Response
-from pydantic import BaseModel, Field
+from pydantic import BaseModel
 
 from backend.data.workspace import (
     WorkspaceFile,
@@ -131,26 +131,9 @@ class StorageUsageResponse(BaseModel):
     file_count: int
 
 
-class WorkspaceFileItem(BaseModel):
-    id: str
-    name: str
-    path: str
-    mime_type: str
-    size_bytes: int
-    metadata: dict = Field(default_factory=dict)
-    created_at: str
-
-
-class ListFilesResponse(BaseModel):
-    files: list[WorkspaceFileItem]
-    offset: int = 0
-    has_more: bool = False
-
-
 @router.get(
     "/files/{file_id}/download",
     summary="Download file by ID",
-    operation_id="getWorkspaceDownloadFileById",
 )
 async def download_file(
     user_id: Annotated[str, fastapi.Security(get_user_id)],
@@ -175,7 +158,6 @@ async def download_file(
 @router.delete(
     "/files/{file_id}",
     summary="Delete a workspace file",
-    operation_id="deleteWorkspaceFile",
 )
 async def delete_workspace_file(
     user_id: Annotated[str, fastapi.Security(get_user_id)],
@@ -201,7 +183,6 @@ async def delete_workspace_file(
 @router.post(
     "/files/upload",
     summary="Upload file to workspace",
-    operation_id="uploadWorkspaceFile",
 )
 async def upload_file(
     user_id: Annotated[str, fastapi.Security(get_user_id)],
@@ -215,9 +196,6 @@ async def upload_file(
     Files are stored in session-scoped paths when session_id is provided,
     so the agent's session-scoped tools can discover them automatically.
     """
-    # Empty-string session_id drops session scoping; normalize to None.
-    session_id = session_id or None
-
     config = Config()
 
     # Sanitize filename — strip any directory components
@@ -272,27 +250,16 @@ async def upload_file(
     manager = WorkspaceManager(user_id, workspace.id, session_id)
     try:
         workspace_file = await manager.write_file(
-            content, filename, overwrite=overwrite, metadata={"origin": "user-upload"}
+            content, filename, overwrite=overwrite
         )
     except ValueError as e:
-        # write_file raises ValueError for both path-conflict and size-limit
-        # cases; map each to its correct HTTP status.
-        message = str(e)
-        if message.startswith("File too large"):
-            raise fastapi.HTTPException(status_code=413, detail=message) from e
-        raise fastapi.HTTPException(status_code=409, detail=message) from e
+        raise fastapi.HTTPException(status_code=409, detail=str(e)) from e
 
     # Post-write storage check — eliminates TOCTOU race on the quota.
     # If a concurrent upload pushed us over the limit, undo this write.
     new_total = await get_workspace_total_size(workspace.id)
     if storage_limit_bytes and new_total > storage_limit_bytes:
-        try:
-            await soft_delete_workspace_file(workspace_file.id, workspace.id)
-        except Exception as e:
-            logger.warning(
-                f"Failed to soft-delete over-quota file {workspace_file.id} "
-                f"in workspace {workspace.id}: {e}"
-            )
+        await soft_delete_workspace_file(workspace_file.id, workspace.id)
         raise fastapi.HTTPException(
             status_code=413,
             detail={
@@ -314,7 +281,6 @@ async def upload_file(
 @router.get(
     "/storage/usage",
     summary="Get workspace storage usage",
-    operation_id="getWorkspaceStorageUsage",
 )
 async def get_storage_usage(
     user_id: Annotated[str, fastapi.Security(get_user_id)],
@@ -335,57 +301,3 @@ async def get_storage_usage(
         used_percent=round((used_bytes / limit_bytes) * 100, 1) if limit_bytes else 0,
         file_count=file_count,
     )
-
-
-@router.get(
-    "/files",
-    summary="List workspace files",
-    operation_id="listWorkspaceFiles",
-)
-async def list_workspace_files(
-    user_id: Annotated[str, fastapi.Security(get_user_id)],
-    session_id: str | None = Query(default=None),
-    limit: int = Query(default=200, ge=1, le=1000),
-    offset: int = Query(default=0, ge=0),
-) -> ListFilesResponse:
-    """
-    List files in the user's workspace.
-
-    When session_id is provided, only files for that session are returned.
-    Otherwise, all files across sessions are listed. Results are paginated
-    via `limit`/`offset`; `has_more` indicates whether additional pages exist.
-    """
-    workspace = await get_or_create_workspace(user_id)
-
-    # Treat empty-string session_id the same as omitted — an empty value
-    # would otherwise silently list files across every session instead of
-    # scoping to one.
-    session_id = session_id or None
-
-    manager = WorkspaceManager(user_id, workspace.id, session_id)
-    include_all = session_id is None
-    # Fetch one extra to compute has_more without a separate count query.
-    files = await manager.list_files(
-        limit=limit + 1,
-        offset=offset,
-        include_all_sessions=include_all,
-    )
-    has_more = len(files) > limit
-    page = files[:limit]
-
-    return ListFilesResponse(
-        files=[
-            WorkspaceFileItem(
-                id=f.id,
-                name=f.name,
-                path=f.path,
-                mime_type=f.mime_type,
-                size_bytes=f.size_bytes,
-                metadata=f.metadata or {},
-                created_at=f.created_at.isoformat(),
-            )
-            for f in page
-        ],
-        offset=offset,
-        has_more=has_more,
-    )

autogpt_platform/backend/backend/api/features/workspace/routes_test.py

@@ -1,28 +1,48 @@
+"""Tests for workspace file upload and download routes."""
+
 import io
 from datetime import datetime, timezone
-from unittest.mock import AsyncMock, MagicMock, patch
 
 import fastapi
 import fastapi.testclient
 import pytest
+import pytest_mock
 
-from backend.api.features.workspace.routes import router
-from backend.data.workspace import Workspace, WorkspaceFile
+from backend.api.features.workspace import routes as workspace_routes
+from backend.data.workspace import WorkspaceFile
 
 app = fastapi.FastAPI()
-app.include_router(router)
+app.include_router(workspace_routes.router)
 
 
 @app.exception_handler(ValueError)
 async def _value_error_handler(
     request: fastapi.Request, exc: ValueError
 ) -> fastapi.responses.JSONResponse:
-    """Mirror the production ValueError → 400 mapping from the REST app."""
+    """Mirror the production ValueError → 400 mapping from rest_api.py."""
     return fastapi.responses.JSONResponse(status_code=400, content={"detail": str(exc)})
 
 
 client = fastapi.testclient.TestClient(app)
 
+TEST_USER_ID = "3e53486c-cf57-477e-ba2a-cb02dc828e1a"
+
+MOCK_WORKSPACE = type("W", (), {"id": "ws-1"})()
+
+_NOW = datetime(2023, 1, 1, tzinfo=timezone.utc)
+
+MOCK_FILE = WorkspaceFile(
+    id="file-aaa-bbb",
+    workspace_id="ws-1",
+    created_at=_NOW,
+    updated_at=_NOW,
+    name="hello.txt",
+    path="/session/hello.txt",
+    mime_type="text/plain",
+    size_bytes=13,
+    storage_path="local://hello.txt",
+)
+
 
 @pytest.fixture(autouse=True)
 def setup_app_auth(mock_jwt_user):
@@ -33,201 +53,25 @@ def setup_app_auth(mock_jwt_user):
     app.dependency_overrides.clear()
 
 
-def _make_workspace(user_id: str = "test-user-id") -> Workspace:
-    return Workspace(
-        id="ws-001",
-        user_id=user_id,
-        created_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
-        updated_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
-    )
-
-
-def _make_file(**overrides) -> WorkspaceFile:
-    defaults = {
-        "id": "file-001",
-        "workspace_id": "ws-001",
-        "created_at": datetime(2026, 1, 1, tzinfo=timezone.utc),
-        "updated_at": datetime(2026, 1, 1, tzinfo=timezone.utc),
-        "name": "test.txt",
-        "path": "/test.txt",
-        "storage_path": "local://test.txt",
-        "mime_type": "text/plain",
-        "size_bytes": 100,
-        "checksum": None,
-        "is_deleted": False,
-        "deleted_at": None,
-        "metadata": {},
-    }
-    defaults.update(overrides)
-    return WorkspaceFile(**defaults)
-
-
-def _make_file_mock(**overrides) -> MagicMock:
-    """Create a mock WorkspaceFile to simulate DB records with null fields."""
-    defaults = {
-        "id": "file-001",
-        "name": "test.txt",
-        "path": "/test.txt",
-        "mime_type": "text/plain",
-        "size_bytes": 100,
-        "metadata": {},
-        "created_at": datetime(2026, 1, 1, tzinfo=timezone.utc),
-    }
-    defaults.update(overrides)
-    mock = MagicMock(spec=WorkspaceFile)
-    for k, v in defaults.items():
-        setattr(mock, k, v)
-    return mock
-
-
-# -- list_workspace_files tests --
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_returns_all_when_no_session(mock_manager_cls, mock_get_workspace):
-    mock_get_workspace.return_value = _make_workspace()
-    files = [
-        _make_file(id="f1", name="a.txt", metadata={"origin": "user-upload"}),
-        _make_file(id="f2", name="b.csv", metadata={"origin": "agent-created"}),
-    ]
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = files
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files")
-    assert response.status_code == 200
-
-    data = response.json()
-    assert len(data["files"]) == 2
-    assert data["has_more"] is False
-    assert data["offset"] == 0
-    assert data["files"][0]["id"] == "f1"
-    assert data["files"][0]["metadata"] == {"origin": "user-upload"}
-    assert data["files"][1]["id"] == "f2"
-    mock_instance.list_files.assert_called_once_with(
-        limit=201, offset=0, include_all_sessions=True
-    )
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_scopes_to_session_when_provided(
-    mock_manager_cls, mock_get_workspace, test_user_id
-):
-    mock_get_workspace.return_value = _make_workspace(user_id=test_user_id)
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = []
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files?session_id=sess-123")
-    assert response.status_code == 200
-
-    data = response.json()
-    assert data["files"] == []
-    assert data["has_more"] is False
-    mock_manager_cls.assert_called_once_with(test_user_id, "ws-001", "sess-123")
-    mock_instance.list_files.assert_called_once_with(
-        limit=201, offset=0, include_all_sessions=False
-    )
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_null_metadata_coerced_to_empty_dict(
-    mock_manager_cls, mock_get_workspace
-):
-    """Route uses `f.metadata or {}` for pre-existing files with null metadata."""
-    mock_get_workspace.return_value = _make_workspace()
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = [_make_file_mock(metadata=None)]
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files")
-    assert response.status_code == 200
-    assert response.json()["files"][0]["metadata"] == {}
-
-
-# -- upload_file metadata tests --
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.get_workspace_total_size")
-@patch("backend.api.features.workspace.routes.scan_content_safe")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_upload_passes_user_upload_origin_metadata(
-    mock_manager_cls, mock_scan, mock_total_size, mock_get_workspace
-):
-    mock_get_workspace.return_value = _make_workspace()
-    mock_total_size.return_value = 100
-    written = _make_file(id="new-file", name="doc.pdf")
-    mock_instance = AsyncMock()
-    mock_instance.write_file.return_value = written
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.post(
-        "/files/upload",
-        files={"file": ("doc.pdf", b"fake-pdf-content", "application/pdf")},
-    )
-    assert response.status_code == 200
-
-    mock_instance.write_file.assert_called_once()
-    call_kwargs = mock_instance.write_file.call_args
-    assert call_kwargs.kwargs.get("metadata") == {"origin": "user-upload"}
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.get_workspace_total_size")
-@patch("backend.api.features.workspace.routes.scan_content_safe")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_upload_returns_409_on_file_conflict(
-    mock_manager_cls, mock_scan, mock_total_size, mock_get_workspace
-):
-    mock_get_workspace.return_value = _make_workspace()
-    mock_total_size.return_value = 100
-    mock_instance = AsyncMock()
-    mock_instance.write_file.side_effect = ValueError("File already exists at path")
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.post(
-        "/files/upload",
-        files={"file": ("dup.txt", b"content", "text/plain")},
-    )
-    assert response.status_code == 409
-    assert "already exists" in response.json()["detail"]
-
-
-# -- Restored upload/download/delete security + invariant tests --
-
-
 def _upload(
     filename: str = "hello.txt",
     content: bytes = b"Hello, world!",
     content_type: str = "text/plain",
 ):
+    """Helper to POST a file upload."""
     return client.post(
         "/files/upload?session_id=sess-1",
         files={"file": (filename, io.BytesIO(content), content_type)},
     )
 
 
-_MOCK_FILE = WorkspaceFile(
-    id="file-aaa-bbb",
-    workspace_id="ws-001",
-    created_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
-    updated_at=datetime(2026, 1, 1, tzinfo=timezone.utc),
-    name="hello.txt",
-    path="/sessions/sess-1/hello.txt",
-    mime_type="text/plain",
-    size_bytes=13,
-    storage_path="local://hello.txt",
-)
+# ---- Happy path ----
 
 
-def test_upload_happy_path(mocker):
+def test_upload_happy_path(mocker: pytest_mock.MockFixture):
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
@@ -238,7 +82,7 @@ def test_upload_happy_path(mocker):
         return_value=None,
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
@@ -252,7 +96,10 @@ def test_upload_happy_path(mocker):
     assert data["size_bytes"] == 13
 
 
-def test_upload_exceeds_max_file_size(mocker):
+# ---- Per-file size limit ----
+
+
+def test_upload_exceeds_max_file_size(mocker: pytest_mock.MockFixture):
     """Files larger than max_file_size_mb should be rejected with 413."""
     cfg = mocker.patch("backend.api.features.workspace.routes.Config")
     cfg.return_value.max_file_size_mb = 0  # 0 MB → any content is too big
@@ -262,11 +109,15 @@ def test_upload_exceeds_max_file_size(mocker):
     assert response.status_code == 413
 
 
-def test_upload_storage_quota_exceeded(mocker):
+# ---- Storage quota exceeded ----
+
+
+def test_upload_storage_quota_exceeded(mocker: pytest_mock.MockFixture):
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
+    # Current usage already at limit
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
         return_value=500 * 1024 * 1024,
@@ -277,22 +128,27 @@ def test_upload_storage_quota_exceeded(mocker):
     assert "Storage limit exceeded" in response.text
 
 
-def test_upload_post_write_quota_race(mocker):
-    """Concurrent upload tipping over limit after write should soft-delete + 413."""
+# ---- Post-write quota race (B2) ----
+
+
+def test_upload_post_write_quota_race(mocker: pytest_mock.MockFixture):
+    """If a concurrent upload tips the total over the limit after write,
+    the file should be soft-deleted and 413 returned."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
+    # Pre-write check passes (under limit), but post-write check fails
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
-        side_effect=[0, 600 * 1024 * 1024],
+        side_effect=[0, 600 * 1024 * 1024],  # first call OK, second over limit
     )
     mocker.patch(
         "backend.api.features.workspace.routes.scan_content_safe",
         return_value=None,
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
@@ -304,14 +160,17 @@ def test_upload_post_write_quota_race(mocker):
 
     response = _upload()
     assert response.status_code == 413
-    mock_delete.assert_called_once_with("file-aaa-bbb", "ws-001")
+    mock_delete.assert_called_once_with("file-aaa-bbb", "ws-1")
 
 
-def test_upload_any_extension(mocker):
+# ---- Any extension accepted (no allowlist) ----
+
+
+def test_upload_any_extension(mocker: pytest_mock.MockFixture):
     """Any file extension should be accepted — ClamAV is the security layer."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
@@ -322,7 +181,7 @@ def test_upload_any_extension(mocker):
         return_value=None,
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
@@ -332,13 +191,16 @@ def test_upload_any_extension(mocker):
     assert response.status_code == 200
 
 
-def test_upload_blocked_by_virus_scan(mocker):
+# ---- Virus scan rejection ----
+
+
+def test_upload_blocked_by_virus_scan(mocker: pytest_mock.MockFixture):
     """Files flagged by ClamAV should be rejected and never written to storage."""
     from backend.api.features.store.exceptions import VirusDetectedError
 
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
@@ -349,7 +211,7 @@ def test_upload_blocked_by_virus_scan(mocker):
         side_effect=VirusDetectedError("Eicar-Test-Signature"),
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
@@ -357,14 +219,18 @@ def test_upload_blocked_by_virus_scan(mocker):
 
     response = _upload(filename="evil.exe", content=b"X5O!P%@AP...")
     assert response.status_code == 400
+    assert "Virus detected" in response.text
     mock_manager.write_file.assert_not_called()
 
 
-def test_upload_file_without_extension(mocker):
+# ---- No file extension ----
+
+
+def test_upload_file_without_extension(mocker: pytest_mock.MockFixture):
     """Files without an extension should be accepted and stored as-is."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
@@ -375,7 +241,7 @@ def test_upload_file_without_extension(mocker):
         return_value=None,
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
@@ -391,11 +257,14 @@ def test_upload_file_without_extension(mocker):
     assert mock_manager.write_file.call_args[0][1] == "Makefile"
 
 
-def test_upload_strips_path_components(mocker):
+# ---- Filename sanitization (SF5) ----
+
+
+def test_upload_strips_path_components(mocker: pytest_mock.MockFixture):
     """Path-traversal filenames should be reduced to their basename."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_total_size",
@@ -406,23 +275,28 @@ def test_upload_strips_path_components(mocker):
         return_value=None,
     )
     mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(return_value=_MOCK_FILE)
+    mock_manager.write_file = mocker.AsyncMock(return_value=MOCK_FILE)
     mocker.patch(
         "backend.api.features.workspace.routes.WorkspaceManager",
         return_value=mock_manager,
     )
 
+    # Filename with traversal
     _upload(filename="../../etc/passwd.txt")
 
+    # write_file should have been called with just the basename
     mock_manager.write_file.assert_called_once()
     call_args = mock_manager.write_file.call_args
     assert call_args[0][1] == "passwd.txt"
 
 
-def test_download_file_not_found(mocker):
+# ---- Download ----
+
+
+def test_download_file_not_found(mocker: pytest_mock.MockFixture):
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace_file",
@@ -433,11 +307,14 @@ def test_download_file_not_found(mocker):
     assert response.status_code == 404
 
 
-def test_delete_file_success(mocker):
+# ---- Delete ----
+
+
+def test_delete_file_success(mocker: pytest_mock.MockFixture):
     """Deleting an existing file should return {"deleted": true}."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mock_manager = mocker.MagicMock()
     mock_manager.delete_file = mocker.AsyncMock(return_value=True)
@@ -452,11 +329,11 @@ def test_delete_file_success(mocker):
     mock_manager.delete_file.assert_called_once_with("file-aaa-bbb")
 
 
-def test_delete_file_not_found(mocker):
+def test_delete_file_not_found(mocker: pytest_mock.MockFixture):
     """Deleting a non-existent file should return 404."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace",
-        return_value=_make_workspace(),
+        return_value=MOCK_WORKSPACE,
     )
     mock_manager = mocker.MagicMock()
     mock_manager.delete_file = mocker.AsyncMock(return_value=False)
@@ -470,7 +347,7 @@ def test_delete_file_not_found(mocker):
     assert "File not found" in response.text
 
 
-def test_delete_file_no_workspace(mocker):
+def test_delete_file_no_workspace(mocker: pytest_mock.MockFixture):
     """Deleting when user has no workspace should return 404."""
     mocker.patch(
         "backend.api.features.workspace.routes.get_workspace",
@@ -480,123 +357,3 @@ def test_delete_file_no_workspace(mocker):
     response = client.delete("/files/file-aaa-bbb")
     assert response.status_code == 404
     assert "Workspace not found" in response.text
-
-
-def test_upload_write_file_too_large_returns_413(mocker):
-    """write_file raises ValueError("File too large: …") → must map to 413."""
-    mocker.patch(
-        "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.get_workspace_total_size",
-        return_value=0,
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.scan_content_safe",
-        return_value=None,
-    )
-    mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(
-        side_effect=ValueError("File too large: 900 bytes exceeds 1MB limit")
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.WorkspaceManager",
-        return_value=mock_manager,
-    )
-
-    response = _upload()
-    assert response.status_code == 413
-    assert "File too large" in response.text
-
-
-def test_upload_write_file_conflict_returns_409(mocker):
-    """Non-'File too large' ValueErrors from write_file stay as 409."""
-    mocker.patch(
-        "backend.api.features.workspace.routes.get_or_create_workspace",
-        return_value=_make_workspace(),
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.get_workspace_total_size",
-        return_value=0,
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.scan_content_safe",
-        return_value=None,
-    )
-    mock_manager = mocker.MagicMock()
-    mock_manager.write_file = mocker.AsyncMock(
-        side_effect=ValueError("File already exists at path: /sessions/x/a.txt")
-    )
-    mocker.patch(
-        "backend.api.features.workspace.routes.WorkspaceManager",
-        return_value=mock_manager,
-    )
-
-    response = _upload()
-    assert response.status_code == 409
-    assert "already exists" in response.text
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_has_more_true_when_limit_exceeded(
-    mock_manager_cls, mock_get_workspace
-):
-    """The limit+1 fetch trick must flip has_more=True and trim the page."""
-    mock_get_workspace.return_value = _make_workspace()
-    # Backend was asked for limit+1=3, and returned exactly 3 items.
-    files = [
-        _make_file(id="f1", name="a.txt"),
-        _make_file(id="f2", name="b.txt"),
-        _make_file(id="f3", name="c.txt"),
-    ]
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = files
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files?limit=2")
-    assert response.status_code == 200
-    data = response.json()
-    assert data["has_more"] is True
-    assert len(data["files"]) == 2
-    assert data["files"][0]["id"] == "f1"
-    assert data["files"][1]["id"] == "f2"
-    mock_instance.list_files.assert_called_once_with(
-        limit=3, offset=0, include_all_sessions=True
-    )
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_has_more_false_when_exactly_page_size(
-    mock_manager_cls, mock_get_workspace
-):
-    """Exactly `limit` rows means we're on the last page — has_more=False."""
-    mock_get_workspace.return_value = _make_workspace()
-    files = [_make_file(id="f1", name="a.txt"), _make_file(id="f2", name="b.txt")]
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = files
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files?limit=2")
-    assert response.status_code == 200
-    data = response.json()
-    assert data["has_more"] is False
-    assert len(data["files"]) == 2
-
-
-@patch("backend.api.features.workspace.routes.get_or_create_workspace")
-@patch("backend.api.features.workspace.routes.WorkspaceManager")
-def test_list_files_offset_is_echoed_back(mock_manager_cls, mock_get_workspace):
-    mock_get_workspace.return_value = _make_workspace()
-    mock_instance = AsyncMock()
-    mock_instance.list_files.return_value = []
-    mock_manager_cls.return_value = mock_instance
-
-    response = client.get("/files?offset=50&limit=10")
-    assert response.status_code == 200
-    assert response.json()["offset"] == 50
-    mock_instance.list_files.assert_called_once_with(
-        limit=11, offset=50, include_all_sessions=True
-    )

autogpt_platform/backend/backend/api/rest_api.py

@@ -18,6 +18,7 @@ from prisma.errors import PrismaError
 
 import backend.api.features.admin.credit_admin_routes
 import backend.api.features.admin.execution_analytics_routes
+import backend.api.features.admin.platform_cost_routes
 import backend.api.features.admin.rate_limit_admin_routes
 import backend.api.features.admin.store_admin_routes
 import backend.api.features.builder
@@ -329,6 +330,11 @@ app.include_router(
     tags=["v2", "admin"],
     prefix="/api/copilot",
 )
+app.include_router(
+    backend.api.features.admin.platform_cost_routes.router,
+    tags=["v2", "admin"],
+    prefix="/api/platform-costs",
+)
 app.include_router(
     backend.api.features.executions.review.routes.router,
     tags=["v2", "executions", "review"],

autogpt_platform/backend/backend/blocks/ai_shortform_video_block.py

@@ -18,6 +18,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -358,6 +359,7 @@ class AIShortformVideoCreatorBlock(Block):
                 execution_context=execution_context,
                 return_format="for_block_output",
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "video_url", stored_url
 
 
@@ -565,6 +567,7 @@ class AIAdMakerVideoCreatorBlock(Block):
             execution_context=execution_context,
             return_format="for_block_output",
         )
+        self.merge_stats(NodeExecutionStats(output_size=1))
         yield "video_url", stored_url
 
 
@@ -760,4 +763,5 @@ class AIScreenshotToVideoAdBlock(Block):
             execution_context=execution_context,
             return_format="for_block_output",
         )
+        self.merge_stats(NodeExecutionStats(output_size=1))
         yield "video_url", stored_url

autogpt_platform/backend/backend/blocks/apollo/organization.py

@@ -17,7 +17,7 @@ from backend.blocks.apollo.models import (
     PrimaryPhone,
     SearchOrganizationsRequest,
 )
-from backend.data.model import CredentialsField, SchemaField
+from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
 
 
 class SearchOrganizationsBlock(Block):
@@ -218,6 +218,7 @@ To find IDs, identify the values for organization_id when you call this endpoint
     ) -> BlockOutput:
         query = SearchOrganizationsRequest(**input_data.model_dump())
         organizations = await self.search_organizations(query, credentials)
+        self.merge_stats(NodeExecutionStats(output_size=len(organizations)))
         for organization in organizations:
             yield "organization", organization
         yield "organizations", organizations

autogpt_platform/backend/backend/blocks/apollo/people.py

@@ -21,7 +21,7 @@ from backend.blocks.apollo.models import (
     SearchPeopleRequest,
     SenorityLevels,
 )
-from backend.data.model import CredentialsField, SchemaField
+from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
 
 
 class SearchPeopleBlock(Block):
@@ -366,4 +366,5 @@ class SearchPeopleBlock(Block):
                 *(enrich_or_fallback(person) for person in people)
             )
 
+        self.merge_stats(NodeExecutionStats(output_size=len(people)))
         yield "people", people

autogpt_platform/backend/backend/blocks/apollo/person.py

@@ -13,7 +13,7 @@ from backend.blocks.apollo._auth import (
     ApolloCredentialsInput,
 )
 from backend.blocks.apollo.models import Contact, EnrichPersonRequest
-from backend.data.model import CredentialsField, SchemaField
+from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
 
 
 class GetPersonDetailBlock(Block):
@@ -141,4 +141,5 @@ class GetPersonDetailBlock(Block):
         **kwargs,
     ) -> BlockOutput:
         query = EnrichPersonRequest(**input_data.model_dump())
+        self.merge_stats(NodeExecutionStats(output_size=1))
         yield "contact", await self.enrich_person(query, credentials)

autogpt_platform/backend/backend/blocks/code_executor.py

@@ -17,6 +17,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -342,6 +343,7 @@ class ExecuteCodeBlock(Block, BaseE2BExecutorMixin):
 
             # Determine result object shape & filter out empty formats
             main_result, results = self.process_execution_results(results)
+            self.merge_stats(NodeExecutionStats(output_size=1))
             if main_result:
                 yield "main_result", main_result
             yield "results", results
@@ -467,6 +469,7 @@ class InstantiateCodeSandboxBlock(Block, BaseE2BExecutorMixin):
                 setup_commands=input_data.setup_commands,
                 timeout=input_data.timeout,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             if sandbox_id:
                 yield "sandbox_id", sandbox_id
             else:
@@ -577,6 +580,7 @@ class ExecuteCodeStepBlock(Block, BaseE2BExecutorMixin):
 
             # Determine result object shape & filter out empty formats
             main_result, results = self.process_execution_results(results)
+            self.merge_stats(NodeExecutionStats(output_size=1))
             if main_result:
                 yield "main_result", main_result
             yield "results", results

autogpt_platform/backend/backend/blocks/enrichlayer/linkedin.py

@@ -15,7 +15,12 @@ from backend.blocks._base import (
     BlockSchemaInput,
     BlockSchemaOutput,
 )
-from backend.data.model import APIKeyCredentials, CredentialsField, SchemaField
+from backend.data.model import (
+    APIKeyCredentials,
+    CredentialsField,
+    NodeExecutionStats,
+    SchemaField,
+)
 from backend.util.type import MediaFileType
 
 from ._api import (
@@ -195,6 +200,7 @@ class GetLinkedinProfileBlock(Block):
                 include_social_media=input_data.include_social_media,
                 include_extra=input_data.include_extra,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "profile", profile
         except Exception as e:
             logger.error(f"Error fetching LinkedIn profile: {str(e)}")
@@ -341,6 +347,7 @@ class LinkedinPersonLookupBlock(Block):
                 include_similarity_checks=input_data.include_similarity_checks,
                 enrich_profile=input_data.enrich_profile,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "lookup_result", lookup_result
         except Exception as e:
             logger.error(f"Error looking up LinkedIn profile: {str(e)}")
@@ -443,6 +450,7 @@ class LinkedinRoleLookupBlock(Block):
                 company_name=input_data.company_name,
                 enrich_profile=input_data.enrich_profile,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "role_lookup_result", role_lookup_result
         except Exception as e:
             logger.error(f"Error looking up role in company: {str(e)}")
@@ -523,6 +531,7 @@ class GetLinkedinProfilePictureBlock(Block):
                 credentials=credentials,
                 linkedin_profile_url=input_data.linkedin_profile_url,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "profile_picture_url", profile_picture
         except Exception as e:
             logger.error(f"Error getting profile picture: {str(e)}")

autogpt_platform/backend/backend/blocks/exa/contents.py

@@ -4,6 +4,7 @@ from typing import Optional
 from exa_py import AsyncExa
 from pydantic import BaseModel
 
+from backend.data.model import NodeExecutionStats
 from backend.sdk import (
     APIKeyCredentials,
     Block,
@@ -223,3 +224,6 @@ class ExaContentsBlock(Block):
 
         if response.cost_dollars:
             yield "cost_dollars", response.cost_dollars
+            self.merge_stats(
+                NodeExecutionStats(provider_cost=response.cost_dollars.total)
+            )

autogpt_platform/backend/backend/blocks/exa/search.py

@@ -4,6 +4,7 @@ from typing import Optional
 
 from exa_py import AsyncExa
 
+from backend.data.model import NodeExecutionStats
 from backend.sdk import (
     APIKeyCredentials,
     Block,
@@ -206,3 +207,6 @@ class ExaSearchBlock(Block):
 
         if response.cost_dollars:
             yield "cost_dollars", response.cost_dollars
+            self.merge_stats(
+                NodeExecutionStats(provider_cost=response.cost_dollars.total)
+            )

autogpt_platform/backend/backend/blocks/fal/ai_video_generator.py

@@ -18,7 +18,7 @@ from backend.blocks.fal._auth import (
     FalCredentialsInput,
 )
 from backend.data.execution import ExecutionContext
-from backend.data.model import SchemaField
+from backend.data.model import NodeExecutionStats, SchemaField
 from backend.util.file import store_media_file
 from backend.util.request import ClientResponseError, Requests
 from backend.util.type import MediaFileType
@@ -230,6 +230,7 @@ class AIVideoGeneratorBlock(Block):
                 execution_context=execution_context,
                 return_format="for_block_output",
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "video_url", stored_url
         except Exception as e:
             error_message = str(e)

autogpt_platform/backend/backend/blocks/google_maps.py

@@ -14,6 +14,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -117,6 +118,7 @@ class GoogleMapsSearchBlock(Block):
             input_data.radius,
             input_data.max_results,
         )
+        self.merge_stats(NodeExecutionStats(output_size=len(places)))
         for place in places:
             yield "place", place
 

autogpt_platform/backend/backend/blocks/ideogram.py

@@ -14,6 +14,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -227,6 +228,7 @@ class IdeogramModelBlock(Block):
                 image_url=result,
             )
 
+        self.merge_stats(NodeExecutionStats(output_size=1))
         yield "result", result
 
     async def run_model(

autogpt_platform/backend/backend/blocks/jina/embeddings.py

@@ -10,7 +10,7 @@ from backend.blocks.jina._auth import (
     JinaCredentialsField,
     JinaCredentialsInput,
 )
-from backend.data.model import SchemaField
+from backend.data.model import NodeExecutionStats, SchemaField
 from backend.util.request import Requests
 
 
@@ -45,5 +45,13 @@ class JinaEmbeddingBlock(Block):
         }
         data = {"input": input_data.texts, "model": input_data.model}
         response = await Requests().post(url, headers=headers, json=data)
-        embeddings = [e["embedding"] for e in response.json()["data"]]
+        resp_json = response.json()
+        embeddings = [e["embedding"] for e in resp_json["data"]]
+        usage = resp_json.get("usage", {})
+        if usage.get("total_tokens"):
+            self.merge_stats(
+                NodeExecutionStats(
+                    input_token_count=usage.get("total_tokens", 0),
+                )
+            )
         yield "embeddings", embeddings

autogpt_platform/backend/backend/blocks/llm.py

@@ -205,19 +205,6 @@ class LlmModel(str, Enum, metaclass=LlmModelMeta):
     KIMI_K2 = "moonshotai/kimi-k2"
     QWEN3_235B_A22B_THINKING = "qwen/qwen3-235b-a22b-thinking-2507"
     QWEN3_CODER = "qwen/qwen3-coder"
-    # Z.ai (Zhipu) models
-    ZAI_GLM_4_32B = "z-ai/glm-4-32b"
-    ZAI_GLM_4_5 = "z-ai/glm-4.5"
-    ZAI_GLM_4_5_AIR = "z-ai/glm-4.5-air"
-    ZAI_GLM_4_5_AIR_FREE = "z-ai/glm-4.5-air:free"
-    ZAI_GLM_4_5V = "z-ai/glm-4.5v"
-    ZAI_GLM_4_6 = "z-ai/glm-4.6"
-    ZAI_GLM_4_6V = "z-ai/glm-4.6v"
-    ZAI_GLM_4_7 = "z-ai/glm-4.7"
-    ZAI_GLM_4_7_FLASH = "z-ai/glm-4.7-flash"
-    ZAI_GLM_5 = "z-ai/glm-5"
-    ZAI_GLM_5_TURBO = "z-ai/glm-5-turbo"
-    ZAI_GLM_5V_TURBO = "z-ai/glm-5v-turbo"
     # Llama API models
     LLAMA_API_LLAMA_4_SCOUT = "Llama-4-Scout-17B-16E-Instruct-FP8"
     LLAMA_API_LLAMA4_MAVERICK = "Llama-4-Maverick-17B-128E-Instruct-FP8"
@@ -643,43 +630,6 @@ MODEL_METADATA = {
     LlmModel.QWEN3_CODER: ModelMetadata(
         "open_router", 262144, 262144, "Qwen 3 Coder", "OpenRouter", "Qwen", 3
     ),
-    # https://openrouter.ai/models?q=z-ai
-    LlmModel.ZAI_GLM_4_32B: ModelMetadata(
-        "open_router", 128000, 128000, "GLM 4 32B", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_5: ModelMetadata(
-        "open_router", 131072, 98304, "GLM 4.5", "OpenRouter", "Z.ai", 2
-    ),
-    LlmModel.ZAI_GLM_4_5_AIR: ModelMetadata(
-        "open_router", 131072, 98304, "GLM 4.5 Air", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_5_AIR_FREE: ModelMetadata(
-        "open_router", 131072, 96000, "GLM 4.5 Air (Free)", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_5V: ModelMetadata(
-        "open_router", 65536, 16384, "GLM 4.5V", "OpenRouter", "Z.ai", 2
-    ),
-    LlmModel.ZAI_GLM_4_6: ModelMetadata(
-        "open_router", 204800, 204800, "GLM 4.6", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_6V: ModelMetadata(
-        "open_router", 131072, 131072, "GLM 4.6V", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_7: ModelMetadata(
-        "open_router", 202752, 65535, "GLM 4.7", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_4_7_FLASH: ModelMetadata(
-        "open_router", 202752, 202752, "GLM 4.7 Flash", "OpenRouter", "Z.ai", 1
-    ),
-    LlmModel.ZAI_GLM_5: ModelMetadata(
-        "open_router", 80000, 80000, "GLM 5", "OpenRouter", "Z.ai", 2
-    ),
-    LlmModel.ZAI_GLM_5_TURBO: ModelMetadata(
-        "open_router", 202752, 131072, "GLM 5 Turbo", "OpenRouter", "Z.ai", 3
-    ),
-    LlmModel.ZAI_GLM_5V_TURBO: ModelMetadata(
-        "open_router", 202752, 131072, "GLM 5V Turbo", "OpenRouter", "Z.ai", 3
-    ),
     # Llama API models
     LlmModel.LLAMA_API_LLAMA_4_SCOUT: ModelMetadata(
         "llama_api",
@@ -737,6 +687,7 @@ class LLMResponse(BaseModel):
     prompt_tokens: int
     completion_tokens: int
     reasoning: Optional[str] = None
+    provider_cost: float | None = None
 
 
 def convert_openai_tool_fmt_to_anthropic(
@@ -1095,6 +1046,16 @@ async def llm_call(
         tool_calls = extract_openai_tool_calls(response)
         reasoning = extract_openai_reasoning(response)
 
+        cost = None
+        try:
+            raw_resp = getattr(response, "_response", None)
+            if raw_resp and hasattr(raw_resp, "headers"):
+                cost_header = raw_resp.headers.get("x-total-cost")
+                if cost_header:
+                    cost = float(cost_header)
+        except (ValueError, AttributeError):
+            pass
+
         return LLMResponse(
             raw_response=response.choices[0].message,
             prompt=prompt,
@@ -1103,6 +1064,7 @@ async def llm_call(
             prompt_tokens=response.usage.prompt_tokens if response.usage else 0,
             completion_tokens=response.usage.completion_tokens if response.usage else 0,
             reasoning=reasoning,
+            provider_cost=cost,
         )
     elif provider == "llama_api":
         tools_param = tools if tools else openai.NOT_GIVEN
@@ -1427,12 +1389,13 @@ class AIStructuredResponseGeneratorBlock(AIBlockBase):
                     max_tokens=input_data.max_tokens,
                 )
                 response_text = llm_response.response
-                self.merge_stats(
-                    NodeExecutionStats(
-                        input_token_count=llm_response.prompt_tokens,
-                        output_token_count=llm_response.completion_tokens,
-                    )
+                cost_stats = NodeExecutionStats(
+                    input_token_count=llm_response.prompt_tokens,
+                    output_token_count=llm_response.completion_tokens,
                 )
+                if llm_response.provider_cost is not None:
+                    cost_stats.provider_cost = llm_response.provider_cost
+                self.merge_stats(cost_stats)
                 logger.debug(f"LLM attempt-{retry_count} response: {response_text}")
 
                 if input_data.expected_format:

autogpt_platform/backend/backend/blocks/mem0.py

@@ -8,6 +8,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -153,6 +154,7 @@ class AddMemoryBlock(Block, Mem0Base):
                 messages,
                 **params,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
 
             results = result.get("results", [])
             yield "results", results
@@ -255,6 +257,7 @@ class SearchMemoryBlock(Block, Mem0Base):
             result: list[dict[str, Any]] = client.search(
                 input_data.query, version="v2", filters=filters
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "memories", result
 
         except Exception as e:
@@ -340,6 +343,7 @@ class GetAllMemoriesBlock(Block, Mem0Base):
                 filters=filters,
                 version="v2",
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
 
             yield "memories", memories
 
@@ -434,6 +438,7 @@ class GetLatestMemoryBlock(Block, Mem0Base):
                 filters=filters,
                 version="v2",
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
 
             if memories:
                 # Return the latest memory (first in the list as they're sorted by recency)

autogpt_platform/backend/backend/blocks/nvidia/deepfake.py

@@ -10,7 +10,7 @@ from backend.blocks.nvidia._auth import (
     NvidiaCredentialsField,
     NvidiaCredentialsInput,
 )
-from backend.data.model import SchemaField
+from backend.data.model import NodeExecutionStats, SchemaField
 from backend.util.request import Requests
 from backend.util.type import MediaFileType
 
@@ -69,6 +69,7 @@ class NvidiaDeepfakeDetectBlock(Block):
             data = response.json()
 
             result = data.get("data", [{}])[0]
+            self.merge_stats(NodeExecutionStats(output_size=1))
 
             # Get deepfake probability from first bounding box if any
             deepfake_prob = 0.0

autogpt_platform/backend/backend/blocks/replicate/replicate_block.py

@@ -17,7 +17,12 @@ from backend.blocks.replicate._auth import (
     ReplicateCredentialsInput,
 )
 from backend.blocks.replicate._helper import ReplicateOutputs, extract_result
-from backend.data.model import APIKeyCredentials, CredentialsField, SchemaField
+from backend.data.model import (
+    APIKeyCredentials,
+    CredentialsField,
+    NodeExecutionStats,
+    SchemaField,
+)
 from backend.util.exceptions import BlockExecutionError, BlockInputError
 
 logger = logging.getLogger(__name__)
@@ -108,6 +113,7 @@ class ReplicateModelBlock(Block):
             result = await self.run_model(
                 model_ref, input_data.model_inputs, credentials.api_key
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "result", result
             yield "status", "succeeded"
             yield "model_name", input_data.model_name

autogpt_platform/backend/backend/blocks/screenshotone.py

@@ -16,6 +16,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -185,6 +186,7 @@ class ScreenshotWebPageBlock(Block):
                 block_chats=input_data.block_chats,
                 cache=input_data.cache,
             )
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "image", screenshot_data["image"]
         except Exception as e:
             yield "error", str(e)

autogpt_platform/backend/backend/blocks/search.py

@@ -15,6 +15,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -146,6 +147,7 @@ class GetWeatherInformationBlock(Block, GetRequest):
         weather_data = await self.get_request(url, json=True)
 
         if "main" in weather_data and "weather" in weather_data:
+            self.merge_stats(NodeExecutionStats(output_size=1))
             yield "temperature", str(weather_data["main"]["temp"])
             yield "humidity", str(weather_data["main"]["humidity"])
             yield "condition", weather_data["weather"][0]["description"]

autogpt_platform/backend/backend/blocks/smartlead/campaign.py

@@ -23,7 +23,7 @@ from backend.blocks.smartlead.models import (
     SaveSequencesResponse,
     Sequence,
 )
-from backend.data.model import CredentialsField, SchemaField
+from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
 
 
 class CreateCampaignBlock(Block):
@@ -100,6 +100,7 @@ class CreateCampaignBlock(Block):
         **kwargs,
     ) -> BlockOutput:
         response = await self.create_campaign(input_data.name, credentials)
+        self.merge_stats(NodeExecutionStats(output_size=1))
 
         yield "id", response.id
         yield "name", response.name
@@ -226,6 +227,7 @@ class AddLeadToCampaignBlock(Block):
         response = await self.add_leads_to_campaign(
             input_data.campaign_id, input_data.lead_list, credentials
         )
+        self.merge_stats(NodeExecutionStats(output_size=len(input_data.lead_list)))
 
         yield "campaign_id", input_data.campaign_id
         yield "upload_count", response.upload_count
@@ -321,6 +323,7 @@ class SaveCampaignSequencesBlock(Block):
         response = await self.save_campaign_sequences(
             input_data.campaign_id, input_data.sequences, credentials
         )
+        self.merge_stats(NodeExecutionStats(output_size=1))
 
         if response.data:
             yield "data", response.data

autogpt_platform/backend/backend/blocks/sql_query_block.py

@@ -66,11 +66,7 @@ class SQLQueryBlock(Block):
             advanced=False,
         )
         host: SecretStr = SchemaField(
-            description=(
-                "Database hostname or IP address. "
-                "Treated as a secret to avoid leaking infrastructure details. "
-                "Private/internal IPs are blocked (SSRF protection)."
-            ),
+            description="Database hostname or IP address",
             placeholder="db.example.com",
             secret=True,
         )
@@ -121,12 +117,6 @@ class SQLQueryBlock(Block):
             description="Column names from the query result"
         )
         row_count: int = SchemaField(description="Number of rows returned")
-        truncated: bool = SchemaField(
-            description=(
-                "True when the result set was capped by max_rows, "
-                "indicating additional rows exist in the database"
-            )
-        )
         affected_rows: int = SchemaField(
             description="Number of rows affected by a write query (INSERT/UPDATE/DELETE)"
         )
@@ -157,14 +147,12 @@ class SQLQueryBlock(Block):
                 ("results", [{"test_col": 1}]),
                 ("columns", ["test_col"]),
                 ("row_count", 1),
-                ("truncated", False),
             ],
             test_mock={
                 "execute_query": lambda *_args, **_kwargs: (
                     [{"test_col": 1}],
                     ["test_col"],
                     -1,
-                    False,
                 ),
                 "check_host_allowed": lambda *_args, **_kwargs: ["127.0.0.1"],
             },
@@ -189,8 +177,8 @@ class SQLQueryBlock(Block):
         max_rows: int,
         read_only: bool = True,
         database_type: DatabaseType = DatabaseType.POSTGRES,
-    ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
-        """Execute a SQL query and return (rows, columns, affected_rows, truncated).
+    ) -> tuple[list[dict[str, Any]], list[str], int]:
+        """Execute a SQL query and return (rows, columns, affected_rows).
 
         Delegates to ``_execute_query`` in ``sql_query_helpers``.
         Extracted as a method so it can be mocked during block tests.
@@ -249,7 +237,7 @@ class SQLQueryBlock(Block):
             )
 
         try:
-            results, columns, affected, truncated = await asyncio.to_thread(
+            results, columns, affected = await asyncio.to_thread(
                 self.execute_query,
                 connection_url=connection_url,
                 query=input_data.query,
@@ -261,29 +249,22 @@ class SQLQueryBlock(Block):
             yield "results", results
             yield "columns", columns
             yield "row_count", len(results)
-            yield "truncated", truncated
             if affected >= 0:
                 yield "affected_rows", affected
         except OperationalError as e:
-            yield (
-                "error",
-                self._classify_operational_error(
-                    _sanitize(e),
-                    input_data.timeout,
-                ),
+            yield "error", self._classify_operational_error(
+                _sanitize(e),
+                input_data.timeout,
             )
         except ProgrammingError as e:
             yield "error", f"SQL error: {_sanitize(e)}"
         except DBAPIError as e:
             yield "error", f"Database error: {_sanitize(e)}"
         except ModuleNotFoundError:
-            yield (
-                "error",
-                (
-                    f"Database driver not available for "
-                    f"{input_data.database_type.value}. "
-                    f"Please contact the platform administrator."
-                ),
+            yield "error", (
+                f"Database driver not available for "
+                f"{input_data.database_type.value}. "
+                f"Please contact the platform administrator."
             )
 
     @staticmethod

autogpt_platform/backend/backend/blocks/sql_query_block_test.py

@@ -4,20 +4,15 @@ error sanitization, SSRF protection, and read/write mode behavior."""
 from datetime import date, datetime, time
 from decimal import Decimal
 from typing import Any
-from unittest.mock import AsyncMock, MagicMock, patch
+from unittest.mock import AsyncMock
 
 import pytest
 from pydantic import SecretStr
-from sqlalchemy import create_engine, text
 from sqlalchemy.exc import OperationalError
 
 from backend.blocks.sql_query_block import SQLQueryBlock, UserPasswordCredentials
 from backend.blocks.sql_query_helpers import (
-    _CONNECT_TIMEOUT_SECONDS,
     DatabaseType,
-    _configure_session,
-    _execute_query,
-    _run_in_transaction,
     _sanitize_error,
     _serialize_value,
     _validate_query_is_read_only,
@@ -563,12 +558,11 @@ class TestSQLQueryBlockRunErrorHandling:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         mock_rows = [{"id": 1, "name": "Alice"}, {"id": 2, "name": "Bob"}]
         mock_cols = ["id", "name"]
-        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert outputs["results"] == mock_rows
         assert outputs["columns"] == mock_cols
         assert outputs["row_count"] == 2
-        assert outputs["truncated"] is False
         # SELECT does not produce affected_rows
         assert "affected_rows" not in outputs
 
@@ -591,7 +585,7 @@ class TestSQLQueryBlockRunErrorHandling:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         mock_rows = [{"id": 1}]
         mock_cols = ["id"]
-        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["results"] == mock_rows
@@ -616,7 +610,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["results"] == []
@@ -633,7 +627,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["affected_rows"] == 1
@@ -648,7 +642,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["affected_rows"] == 1
@@ -663,7 +657,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 0, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 0)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
 
@@ -677,7 +671,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 0, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 0)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
 
@@ -721,7 +715,7 @@ class TestSQLQueryBlockWriteMode:
             read_only=False,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([], [], 42, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([], [], 42)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["affected_rows"] == 42
@@ -738,7 +732,7 @@ class TestSQLQueryBlockWriteMode:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         mock_rows = [{"id": 1, "name": "Alice"}]
         mock_cols = ["id", "name"]
-        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: (mock_rows, mock_cols, -1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert outputs["results"] == mock_rows
         assert outputs["columns"] == mock_cols
@@ -766,7 +760,7 @@ class TestSQLQueryBlockReadOnlyMode:
             read_only=True,
         )
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-        block.execute_query = lambda **_kwargs: ([{"col": 1}], ["col"], -1, False)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: ([{"col": 1}], ["col"], -1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["results"] == [{"col": 1}]
@@ -829,11 +823,9 @@ class TestSQLQueryBlockDefaultPort:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_conn_str = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_conn_str["value"] = str(kwargs["connection_url"])
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -854,11 +846,9 @@ class TestSQLQueryBlockDefaultPort:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_conn_str = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_conn_str["value"] = str(kwargs["connection_url"])
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -879,11 +869,9 @@ class TestSQLQueryBlockDefaultPort:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_conn_str = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_conn_str["value"] = str(kwargs["connection_url"])
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -904,11 +892,9 @@ class TestSQLQueryBlockDefaultPort:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_conn_str = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_conn_str["value"] = str(kwargs["connection_url"])
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -933,11 +919,9 @@ class TestSQLQueryBlockDNSPinning:
         block.check_host_allowed = AsyncMock(return_value=["93.184.216.34"])  # type: ignore[assignment]
         captured_conn_str = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_conn_str["value"] = str(kwargs["connection_url"])
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -1074,11 +1058,9 @@ class TestTimeoutEnforcement:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_kwargs: dict[str, Any] = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_kwargs.update(kwargs)
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -1118,11 +1100,9 @@ class TestMaxRowsEnforcement:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured_kwargs: dict[str, Any] = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured_kwargs.update(kwargs)
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -1141,12 +1121,11 @@ class TestMaxRowsEnforcement:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         # Simulate the database returning exactly max_rows rows (truncated)
         mock_rows = [{"id": i} for i in range(max_rows)]
-        block.execute_query = lambda **_kwargs: (mock_rows, ["id"], -1, True)  # type: ignore[assignment]
+        block.execute_query = lambda **_kwargs: (mock_rows, ["id"], -1)  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
         assert "error" not in outputs
         assert outputs["row_count"] == max_rows
         assert len(outputs["results"]) == max_rows
-        assert outputs["truncated"] is True
 
 
 class TestPasswordInErrorMessages:
@@ -1403,11 +1382,9 @@ class TestURLCreateSpecialCharacters:
         block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
         captured: dict[str, Any] = {}
 
-        def fake_execute(
-            **kwargs: Any,
-        ) -> tuple[list[dict[str, Any]], list[str], int, bool]:
+        def fake_execute(**kwargs: Any) -> tuple[list[dict[str, Any]], list[str], int]:
             captured.update(kwargs)
-            return [{"id": 1}], ["id"], -1, False
+            return [{"id": 1}], ["id"], -1
 
         block.execute_query = fake_execute  # type: ignore[assignment]
         outputs = await _collect_outputs(block, input_data, creds)
@@ -1417,435 +1394,3 @@ class TestURLCreateSpecialCharacters:
         assert hasattr(conn_url, "password"), "connection_url should be a URL object"
         assert conn_url.password == special_pass
         assert conn_url.username == special_user
-
-
-# ---------------------------------------------------------------------------
-# Tests for newly-added disallowed keywords (LOAD, REPLACE, MERGE, BULK, EXEC)
-# ---------------------------------------------------------------------------
-
-
-class TestNewDisallowedKeywords:
-    """LOAD, REPLACE, MERGE, BULK, and EXEC must be blocked in read-only mode."""
-
-    @pytest.mark.parametrize(
-        "query,expected_keyword",
-        [
-            # MySQL file exfiltration
-            (
-                "SELECT * FROM (LOAD DATA LOCAL INFILE '/etc/passwd' INTO TABLE t) AS x",
-                "LOAD",
-            ),
-            # MySQL REPLACE (INSERT-or-UPDATE)
-            (
-                "SELECT * FROM (REPLACE INTO t VALUES (1, 'a')) AS x",
-                "REPLACE",
-            ),
-            # ANSI MERGE (UPSERT)
-            (
-                "SELECT * FROM (MERGE INTO t USING s ON t.id = s.id WHEN MATCHED THEN UPDATE SET t.v = s.v) AS x",
-                "MERGE",
-            ),
-            # MSSQL BULK INSERT -- sqlparse may match INSERT before BULK,
-            # but the query is still blocked by the disallowed keyword check.
-            (
-                "SELECT * FROM (BULK INSERT t FROM '/data.csv') AS x",
-                "BULK",
-            ),
-            # MSSQL EXEC stored procedure
-            (
-                "SELECT * FROM (EXEC sp_who2) AS x",
-                "EXEC",
-            ),
-        ],
-    )
-    def test_keyword_blocked_in_subquery_read_only(
-        self, query: str, expected_keyword: str
-    ):
-        """If sqlparse parses the statement as SELECT, the keyword check
-        must still catch the disallowed keyword inside it."""
-        error, stmt = _validate_single_statement(query)
-        if error is None and stmt is not None:
-            ro_error = _validate_query_is_read_only(stmt)
-            assert (
-                ro_error is not None
-            ), f"Expected '{expected_keyword}' to be blocked but query passed: {query}"
-            assert "Disallowed SQL keyword" in ro_error
-
-    @pytest.mark.parametrize(
-        "query",
-        [
-            "LOAD DATA LOCAL INFILE '/etc/passwd' INTO TABLE t",
-            "REPLACE INTO users (id, name) VALUES (1, 'Alice')",
-            "MERGE INTO target USING source ON target.id = source.id "
-            "WHEN MATCHED THEN UPDATE SET target.name = source.name",
-            "BULK INSERT my_table FROM '/tmp/data.csv'",
-            "EXEC sp_who2",
-        ],
-    )
-    def test_standalone_keyword_rejected_as_non_select(self, query: str):
-        """Standalone statements using these keywords are rejected because
-        their statement type is not SELECT."""
-        error, stmt = _validate_single_statement(query)
-        if error is None and stmt is not None:
-            ro_error = _validate_query_is_read_only(stmt)
-            assert ro_error is not None
-
-    def test_bare_keyword_column_names_blocked(self):
-        """Bare column names that match disallowed keywords (load, replace,
-        merge) ARE blocked because sqlparse classifies them as Keyword tokens.
-        Users should quote such column names (e.g. "load") to avoid this."""
-        query = "SELECT load, replace, merge FROM metrics"
-        error, stmt = _validate_single_statement(query)
-        assert error is None
-        assert stmt is not None
-        ro_error = _validate_query_is_read_only(stmt)
-        assert ro_error is not None
-
-    def test_quoted_keyword_column_names_allowed(self):
-        """Double-quoted column names that happen to match disallowed keywords
-        must NOT trigger false positives (sqlparse classifies them as Names)."""
-        query = 'SELECT "load", "replace", "merge" FROM metrics'
-        error, stmt = _validate_single_statement(query)
-        assert error is None
-        assert stmt is not None
-        ro_error = _validate_query_is_read_only(stmt)
-        assert ro_error is None
-
-    def test_load_in_string_literal_allowed(self):
-        """String literal containing 'LOAD' must not be blocked."""
-        query = "SELECT * FROM logs WHERE message = 'LOAD DATA INFILE test'"
-        error, stmt = _validate_single_statement(query)
-        assert error is None
-        assert stmt is not None
-        ro_error = _validate_query_is_read_only(stmt)
-        assert ro_error is None
-
-
-# ---------------------------------------------------------------------------
-# Decimal NaN / Infinity serialization tests
-# ---------------------------------------------------------------------------
-
-
-class TestSerializeDecimalSpecialValues:
-    """Decimal NaN and Infinity must be serialized as strings, not crash."""
-
-    def test_decimal_nan(self):
-        result = _serialize_value(Decimal("NaN"))
-        assert isinstance(result, str)
-        assert "NaN" in result
-
-    def test_decimal_infinity(self):
-        result = _serialize_value(Decimal("Infinity"))
-        assert isinstance(result, str)
-        assert "Infinity" in result
-
-    def test_decimal_negative_infinity(self):
-        result = _serialize_value(Decimal("-Infinity"))
-        assert isinstance(result, str)
-        assert "-Infinity" in result
-
-    def test_decimal_snan(self):
-        result = _serialize_value(Decimal("sNaN"))
-        assert isinstance(result, str)
-
-
-# ---------------------------------------------------------------------------
-# ModuleNotFoundError path test
-# ---------------------------------------------------------------------------
-
-
-@pytest.mark.asyncio
-class TestModuleNotFoundError:
-    """The block must yield a clean error when a database driver is missing."""
-
-    async def test_missing_driver_yields_clean_error(self):
-        block = SQLQueryBlock()
-        creds = _make_credentials()
-        input_data = _make_input(
-            creds,
-            query="SELECT 1",
-            database_type=DatabaseType.MSSQL,
-            read_only=False,
-        )
-        block.check_host_allowed = AsyncMock(return_value=["1.2.3.4"])  # type: ignore[assignment]
-
-        def raise_module_not_found(**_kwargs: Any) -> None:
-            raise ModuleNotFoundError("No module named 'pymssql'")
-
-        block.execute_query = raise_module_not_found  # type: ignore[assignment]
-        outputs = await _collect_outputs(block, input_data, creds)
-        assert "error" in outputs
-        assert "driver not available" in outputs["error"].lower()
-        assert "mssql" in outputs["error"].lower()
-
-
-# ---------------------------------------------------------------------------
-# Integration tests: _execute_query, _configure_session, _run_in_transaction
-# using an in-memory SQLite database.
-#
-# SQLite does not support SET commands, so _configure_session is a no-op for
-# the "sqlite" dialect.  These tests validate the execution pipeline itself:
-# transaction management, fetchmany truncation, COMMIT vs ROLLBACK, and the
-# new truncated flag.
-# ---------------------------------------------------------------------------
-
-
-def _extract_sql_texts(conn_mock: MagicMock) -> list[str]:
-    """Extract the raw SQL strings from a mocked connection's execute calls."""
-    return [c[0][0].text for c in conn_mock.execute.call_args_list]
-
-
-class TestConfigureSessionDialects:
-    """Verify _configure_session emits correct SET commands per dialect."""
-
-    def test_postgresql_sets_timeout_and_read_only(self):
-        conn = MagicMock()
-        _configure_session(conn, "postgresql", "30000", read_only=True)
-        sqls = _extract_sql_texts(conn)
-        assert any("statement_timeout" in s and "30000" in s for s in sqls)
-        assert any("default_transaction_read_only" in s for s in sqls)
-
-    def test_postgresql_timeout_only_when_not_read_only(self):
-        conn = MagicMock()
-        _configure_session(conn, "postgresql", "5000", read_only=False)
-        assert conn.execute.call_count == 1
-        sqls = _extract_sql_texts(conn)
-        assert "statement_timeout" in sqls[0]
-
-    def test_mysql_sets_max_execution_time_and_read_only(self):
-        conn = MagicMock()
-        _configure_session(conn, "mysql", "10000", read_only=True)
-        sqls = _extract_sql_texts(conn)
-        assert any("MAX_EXECUTION_TIME" in s and "10000" in s for s in sqls)
-        assert any("READ ONLY" in s for s in sqls)
-
-    def test_mysql_timeout_only_when_not_read_only(self):
-        conn = MagicMock()
-        _configure_session(conn, "mysql", "10000", read_only=False)
-        assert conn.execute.call_count == 1
-        sqls = _extract_sql_texts(conn)
-        assert "MAX_EXECUTION_TIME" in sqls[0]
-
-    def test_mssql_sets_lock_timeout(self):
-        conn = MagicMock()
-        _configure_session(conn, "mssql", "15000", read_only=True)
-        sqls = _extract_sql_texts(conn)
-        assert any("LOCK_TIMEOUT" in s and "15000" in s for s in sqls)
-        assert conn.execute.call_count == 1
-
-    def test_unknown_dialect_is_noop(self):
-        conn = MagicMock()
-        _configure_session(conn, "sqlite", "5000", read_only=True)
-        conn.execute.assert_not_called()
-
-
-class TestRunInTransactionSQLite:
-    """Integration tests for _run_in_transaction using a real SQLite engine."""
-
-    def _make_sqlite_engine(self):
-        return create_engine("sqlite:///:memory:")
-
-    def test_select_returns_rows(self):
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER, name TEXT)"))
-            conn.execute(text("INSERT INTO t VALUES (1, 'Alice')"))
-            conn.execute(text("INSERT INTO t VALUES (2, 'Bob')"))
-            results, columns, affected, truncated = _run_in_transaction(
-                conn,
-                "sqlite",
-                "SELECT * FROM t ORDER BY id",
-                max_rows=100,
-                read_only=True,
-            )
-        engine.dispose()
-        assert columns == ["id", "name"]
-        assert len(results) == 2
-        assert results[0] == {"id": 1, "name": "Alice"}
-        assert results[1] == {"id": 2, "name": "Bob"}
-        assert affected == -1
-        assert truncated is False
-
-    def test_fetchmany_truncation(self):
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            for i in range(10):
-                conn.execute(text(f"INSERT INTO t VALUES ({i})"))
-            results, columns, affected, truncated = _run_in_transaction(
-                conn,
-                "sqlite",
-                "SELECT * FROM t ORDER BY id",
-                max_rows=3,
-                read_only=True,
-            )
-        engine.dispose()
-        assert len(results) == 3
-        assert truncated is True
-        assert results[0]["id"] == 0
-        assert results[2]["id"] == 2
-
-    def test_fetchmany_exact_count_not_truncated(self):
-        """When result count equals max_rows, truncated is True because
-        there *might* be more rows (we cannot know without fetching more)."""
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            for i in range(5):
-                conn.execute(text(f"INSERT INTO t VALUES ({i})"))
-            results, _, _, truncated = _run_in_transaction(
-                conn,
-                "sqlite",
-                "SELECT * FROM t",
-                max_rows=5,
-                read_only=True,
-            )
-        engine.dispose()
-        assert len(results) == 5
-        assert truncated is True
-
-    def test_fewer_rows_than_max_not_truncated(self):
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            conn.execute(text("INSERT INTO t VALUES (1)"))
-            results, _, _, truncated = _run_in_transaction(
-                conn,
-                "sqlite",
-                "SELECT * FROM t",
-                max_rows=100,
-                read_only=True,
-            )
-        engine.dispose()
-        assert len(results) == 1
-        assert truncated is False
-
-    def test_rollback_on_read_only(self):
-        """When read_only=True the transaction is rolled back, so writes are
-        not persisted."""
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            conn.execute(text("INSERT INTO t VALUES (1)"))
-            _run_in_transaction(
-                conn,
-                "sqlite",
-                "INSERT INTO t VALUES (2)",
-                max_rows=100,
-                read_only=True,
-            )
-            # The INSERT should have been rolled back
-            result = conn.execute(text("SELECT COUNT(*) FROM t"))
-            count = result.scalar()
-        engine.dispose()
-        assert count == 1
-
-    def test_commit_on_write_mode(self):
-        """When read_only=False the transaction is committed."""
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            conn.execute(text("INSERT INTO t VALUES (1)"))
-            _run_in_transaction(
-                conn,
-                "sqlite",
-                "INSERT INTO t VALUES (2)",
-                max_rows=100,
-                read_only=False,
-            )
-            result = conn.execute(text("SELECT COUNT(*) FROM t"))
-            count = result.scalar()
-        engine.dispose()
-        assert count == 2
-
-    def test_rollback_on_exception(self):
-        """On query error the transaction is rolled back and the exception
-        re-raised."""
-        engine = self._make_sqlite_engine()
-        with engine.connect() as conn:
-            conn = conn.execution_options(isolation_level="AUTOCOMMIT")
-            conn.execute(text("CREATE TABLE t (id INTEGER)"))
-            with pytest.raises(Exception):
-                _run_in_transaction(
-                    conn,
-                    "sqlite",
-                    "SELECT * FROM nonexistent_table",
-                    max_rows=100,
-                    read_only=True,
-                )
-        engine.dispose()
-
-    def test_mssql_begin_transaction_syntax(self):
-        """MSSQL dialect uses 'BEGIN TRANSACTION' instead of 'BEGIN'."""
-        conn = MagicMock()
-        result_mock = MagicMock()
-        result_mock.returns_rows = True
-        result_mock.keys.return_value = ["id"]
-        result_mock.fetchmany.return_value = [(1,)]
-        result_mock.rowcount = 1
-        conn.execute.return_value = result_mock
-
-        _run_in_transaction(conn, "mssql", "SELECT 1", max_rows=10, read_only=True)
-
-        sqls = _extract_sql_texts(conn)
-        assert sqls[0] == "BEGIN TRANSACTION"
-
-    def test_non_mssql_begin_syntax(self):
-        """Non-MSSQL dialects use 'BEGIN' (not 'BEGIN TRANSACTION')."""
-        conn = MagicMock()
-        result_mock = MagicMock()
-        result_mock.returns_rows = True
-        result_mock.keys.return_value = ["id"]
-        result_mock.fetchmany.return_value = [(1,)]
-        result_mock.rowcount = 1
-        conn.execute.return_value = result_mock
-
-        _run_in_transaction(conn, "postgresql", "SELECT 1", max_rows=10, read_only=True)
-
-        sqls = _extract_sql_texts(conn)
-        assert sqls[0] == "BEGIN"
-
-
-class TestExecuteQueryIntegration:
-    """Integration tests for _execute_query using mocked engine internals.
-
-    Note: SQLite does not accept ``connect_timeout`` as a connect_arg, so
-    we cannot pass ``sqlite:///:memory:`` directly to ``_execute_query``.
-    Instead we mock ``create_engine`` to return a real SQLite engine while
-    bypassing the connect_args validation.
-    """
-
-    def test_connect_timeout_constant_used(self):
-        """The _CONNECT_TIMEOUT_SECONDS constant must be passed to create_engine."""
-        with patch("backend.blocks.sql_query_helpers.create_engine") as mock_ce:
-            mock_engine = MagicMock()
-            mock_conn = MagicMock()
-            mock_result = MagicMock()
-            mock_result.returns_rows = True
-            mock_result.keys.return_value = ["x"]
-            mock_result.fetchmany.return_value = [(1,)]
-            mock_result.rowcount = 1
-            mock_conn.execute.return_value = mock_result
-            mock_conn.execution_options.return_value = mock_conn
-            mock_engine.connect.return_value.__enter__ = lambda _: mock_conn
-            mock_engine.connect.return_value.__exit__ = lambda *_: None
-            mock_engine.dialect.name = "postgresql"
-            mock_ce.return_value = mock_engine
-
-            _execute_query(
-                "postgresql://u:p@h/d",
-                "SELECT 1",
-                timeout=10,
-                max_rows=100,
-                read_only=True,
-            )
-
-            mock_ce.assert_called_once()
-            _, kwargs = mock_ce.call_args
-            assert kwargs["connect_args"]["connect_timeout"] == _CONNECT_TIMEOUT_SECONDS

autogpt_platform/backend/backend/blocks/sql_query_helpers.py

@@ -36,16 +36,6 @@ _DISALLOWED_KEYWORDS = {
     "DISCARD",
     "NOTIFY",
     "DO",
-    # MySQL file exfiltration: LOAD DATA LOCAL INFILE reads server/client files
-    "LOAD",
-    # MySQL REPLACE is INSERT-or-UPDATE; data modification
-    "REPLACE",
-    # ANSI MERGE (UPSERT) modifies data
-    "MERGE",
-    # MSSQL BULK INSERT loads external files into tables
-    "BULK",
-    # MSSQL EXEC / EXEC sp_name runs stored procedures (arbitrary code)
-    "EXEC",
 }
 
 # Map DatabaseType enum values to the expected SQLAlchemy driver prefix.
@@ -55,12 +45,6 @@ _DATABASE_TYPE_TO_DRIVER = {
     DatabaseType.MSSQL: "mssql+pymssql",
 }
 
-# Connection timeout in seconds passed to the DBAPI driver (connect_timeout /
-# login_timeout).  This bounds how long the driver waits to establish a TCP
-# connection to the database server.  It is separate from the per-statement
-# timeout configured via SET commands inside _configure_session().
-_CONNECT_TIMEOUT_SECONDS = 10
-
 # Default ports for each database type.
 _DATABASE_TYPE_DEFAULT_PORT = {
     DatabaseType.POSTGRES: 5432,
@@ -280,9 +264,6 @@ def _validate_single_statement(
 def _serialize_value(value: Any) -> Any:
     """Convert database-specific types to JSON-serializable Python types."""
     if isinstance(value, Decimal):
-        # NaN / Infinity are not valid JSON numbers; serialize as strings.
-        if value.is_nan() or value.is_infinite():
-            return str(value)
         # Use int for whole numbers; use str for fractional to preserve exact
         # precision (float would silently round high-precision analytics values).
         if value == value.to_integral_value():
@@ -303,29 +284,7 @@ def _configure_session(
     timeout_ms: str,
     read_only: bool,
 ) -> None:
-    """Set session-level timeout and read-only mode for the given dialect.
-
-    Timeout limitations by database:
-
-    * **PostgreSQL** – ``statement_timeout`` reliably cancels any running
-      statement (SELECT or DML) after the configured duration.
-    * **MySQL** – ``MAX_EXECUTION_TIME`` only applies to **read-only SELECT**
-      statements.  DML (INSERT/UPDATE/DELETE) and DDL are *not* bounded by
-      this hint; they rely on the server's ``wait_timeout`` /
-      ``interactive_timeout`` instead.  There is no session-level setting in
-      MySQL that reliably cancels long-running writes.
-    * **MSSQL** – ``SET LOCK_TIMEOUT`` only limits how long the server waits
-      to acquire a **lock**.  CPU-bound queries (e.g. large scans, hash
-      joins) that do not block on locks will *not* be cancelled.  MSSQL has
-      no session-level ``statement_timeout`` equivalent; the closest
-      mechanism is Resource Governor (requires sysadmin configuration) or
-      ``CONTEXT_INFO``-based external monitoring.
-
-    Note: SQLite is not supported by this block.  The ``_configure_session``
-    function is a no-op for unrecognised dialect names, so an SQLite engine
-    would skip all SET commands silently.  The block's ``DatabaseType`` enum
-    intentionally excludes SQLite.
-    """
+    """Set session-level timeout and read-only mode for the given dialect."""
     if dialect_name == "postgresql":
         conn.execute(text("SET statement_timeout = " + timeout_ms))
         if read_only:
@@ -334,14 +293,13 @@ def _configure_session(
         # NOTE: MAX_EXECUTION_TIME only applies to SELECT statements.
         # Write queries (INSERT/UPDATE/DELETE) are not bounded by this
         # setting; they rely on the database's wait_timeout instead.
-        # See docstring above for full limitations.
         conn.execute(text("SET SESSION MAX_EXECUTION_TIME = " + timeout_ms))
         if read_only:
             conn.execute(text("SET SESSION TRANSACTION READ ONLY"))
     elif dialect_name == "mssql":
-        # MSSQL: SET LOCK_TIMEOUT limits lock-wait time (ms) only.
-        # CPU-bound queries without lock contention are NOT cancelled.
-        # See docstring above for full limitations.
+        # MSSQL: SET LOCK_TIMEOUT limits lock-wait time (ms).
+        # pymssql's connect_args "login_timeout" handles the connection
+        # timeout, but LOCK_TIMEOUT covers in-query lock waits.
         conn.execute(text("SET LOCK_TIMEOUT " + timeout_ms))
         # MSSQL lacks a session-level read-only mode like
         # PostgreSQL/MySQL.  Read-only enforcement is handled by
@@ -355,13 +313,8 @@ def _run_in_transaction(
     query: str,
     max_rows: int,
     read_only: bool,
-) -> tuple[list[dict[str, Any]], list[str], int, bool]:
-    """Execute a query inside an explicit transaction, returning results.
-
-    Returns ``(rows, columns, affected_rows, truncated)`` where *truncated*
-    is ``True`` when ``fetchmany`` returned exactly ``max_rows`` rows,
-    indicating that additional rows may exist in the result set.
-    """
+) -> tuple[list[dict[str, Any]], list[str], int]:
+    """Execute a query inside an explicit transaction, returning results."""
     # MSSQL uses T-SQL "BEGIN TRANSACTION"; others use "BEGIN".
     begin_stmt = "BEGIN TRANSACTION" if dialect_name == "mssql" else "BEGIN"
     conn.execute(text(begin_stmt))
@@ -370,20 +323,16 @@ def _run_in_transaction(
         affected = result.rowcount if not result.returns_rows else -1
         columns = list(result.keys()) if result.returns_rows else []
         rows = result.fetchmany(max_rows) if result.returns_rows else []
-        truncated = len(rows) == max_rows
         results = [
             {col: _serialize_value(val) for col, val in zip(columns, row)}
             for row in rows
         ]
     except Exception:
-        try:
-            conn.execute(text("ROLLBACK"))
-        except Exception:
-            pass
+        conn.execute(text("ROLLBACK"))
         raise
     else:
         conn.execute(text("ROLLBACK" if read_only else "COMMIT"))
-    return results, columns, affected, truncated
+    return results, columns, affected
 
 
 def _execute_query(
@@ -393,12 +342,11 @@ def _execute_query(
     max_rows: int,
     read_only: bool = True,
     database_type: DatabaseType = DatabaseType.POSTGRES,
-) -> tuple[list[dict[str, Any]], list[str], int, bool]:
-    """Execute a SQL query and return (rows, columns, affected_rows, truncated).
+) -> tuple[list[dict[str, Any]], list[str], int]:
+    """Execute a SQL query and return (rows, columns, affected_rows).
 
     Uses SQLAlchemy to connect to any supported database.
     For SELECT queries, rows are limited to ``max_rows`` via DBAPI fetchmany.
-    ``truncated`` is ``True`` when the result set was capped by ``max_rows``.
     For write queries, affected_rows contains the rowcount from the driver.
     When ``read_only`` is True, the database session is set to read-only
     mode and the transaction is always rolled back.
@@ -408,9 +356,7 @@ def _execute_query(
     timeout_key = (
         "login_timeout" if database_type == DatabaseType.MSSQL else "connect_timeout"
     )
-    engine = create_engine(
-        connection_url, connect_args={timeout_key: _CONNECT_TIMEOUT_SECONDS}
-    )
+    engine = create_engine(connection_url, connect_args={timeout_key: 10})
     try:
         with engine.connect() as conn:
             # Use AUTOCOMMIT so SET commands take effect immediately.

autogpt_platform/backend/backend/blocks/talking_head.py

@@ -15,6 +15,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -181,6 +182,7 @@ class CreateTalkingAvatarVideoBlock(Block):
                     execution_context=execution_context,
                     return_format="for_block_output",
                 )
+                self.merge_stats(NodeExecutionStats(output_size=1))
                 yield "video_url", stored_url
                 return
             elif status_response["status"] == "error":

autogpt_platform/backend/backend/blocks/text_to_speech_block.py

@@ -13,6 +13,7 @@ from backend.data.model import (
     APIKeyCredentials,
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
@@ -104,4 +105,5 @@ class UnrealTextToSpeechBlock(Block):
             input_data.text,
             input_data.voice_id,
         )
+        self.merge_stats(NodeExecutionStats(output_size=len(input_data.text)))
         yield "mp3_url", api_response["OutputUri"]

autogpt_platform/backend/backend/blocks/youtube.py

@@ -19,6 +19,7 @@ from backend.blocks._base import (
 from backend.data.model import (
     CredentialsField,
     CredentialsMetaInput,
+    NodeExecutionStats,
     SchemaField,
     UserPasswordCredentials,
 )
@@ -170,6 +171,7 @@ class TranscribeYoutubeVideoBlock(Block):
             transcript = self.get_transcript(video_id, credentials)
             transcript_text = self.format_transcript(transcript=transcript)
 
+            self.merge_stats(NodeExecutionStats(output_size=1))
             # Only yield after all operations succeed
             yield "video_id", video_id
             yield "transcript", transcript_text

autogpt_platform/backend/backend/blocks/zerobounce/validate_emails.py

@@ -21,7 +21,7 @@ from backend.blocks.zerobounce._auth import (
     ZeroBounceCredentials,
     ZeroBounceCredentialsInput,
 )
-from backend.data.model import CredentialsField, SchemaField
+from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
 
 
 class Response(BaseModel):
@@ -177,5 +177,6 @@ class ValidateEmailsBlock(Block):
         )
 
         response_model = Response(**response.__dict__)
+        self.merge_stats(NodeExecutionStats(output_size=1))
 
         yield "response", response_model

autogpt_platform/backend/backend/copilot/baseline/service_unit_test.py

@@ -1,633 +0,0 @@
-"""Unit tests for baseline service pure-logic helpers.
-
-These tests cover ``_baseline_conversation_updater`` and ``_BaselineStreamState``
-without requiring API keys, database connections, or network access.
-"""
-
-from unittest.mock import AsyncMock, patch
-
-import pytest
-from openai.types.chat import ChatCompletionToolParam
-
-from backend.copilot.baseline.service import (
-    _baseline_conversation_updater,
-    _BaselineStreamState,
-    _compress_session_messages,
-    _ThinkingStripper,
-)
-from backend.copilot.model import ChatMessage
-from backend.copilot.transcript_builder import TranscriptBuilder
-from backend.util.prompt import CompressResult
-from backend.util.tool_call_loop import LLMLoopResponse, LLMToolCall, ToolCallResult
-
-
-class TestBaselineStreamState:
-    def test_defaults(self):
-        state = _BaselineStreamState()
-        assert state.pending_events == []
-        assert state.assistant_text == ""
-        assert state.text_started is False
-        assert state.turn_prompt_tokens == 0
-        assert state.turn_completion_tokens == 0
-        assert state.text_block_id  # Should be a UUID string
-
-    def test_mutable_fields(self):
-        state = _BaselineStreamState()
-        state.assistant_text = "hello"
-        state.turn_prompt_tokens = 100
-        state.turn_completion_tokens = 50
-        assert state.assistant_text == "hello"
-        assert state.turn_prompt_tokens == 100
-        assert state.turn_completion_tokens == 50
-
-
-class TestBaselineConversationUpdater:
-    """Tests for _baseline_conversation_updater which updates the OpenAI
-    message list and transcript builder after each LLM call."""
-
-    def _make_transcript_builder(self) -> TranscriptBuilder:
-        builder = TranscriptBuilder()
-        builder.append_user("test question")
-        return builder
-
-    def test_text_only_response(self):
-        """When the LLM returns text without tool calls, the updater appends
-        a single assistant message and records it in the transcript."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text="Hello, world!",
-            tool_calls=[],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=None,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert len(messages) == 1
-        assert messages[0]["role"] == "assistant"
-        assert messages[0]["content"] == "Hello, world!"
-        # Transcript should have user + assistant
-        assert builder.entry_count == 2
-        assert builder.last_entry_type == "assistant"
-
-    def test_tool_calls_response(self):
-        """When the LLM returns tool calls, the updater appends the assistant
-        message with tool_calls and tool result messages."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text="Let me search...",
-            tool_calls=[
-                LLMToolCall(
-                    id="tc_1",
-                    name="search",
-                    arguments='{"query": "test"}',
-                ),
-            ],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-        tool_results = [
-            ToolCallResult(
-                tool_call_id="tc_1",
-                tool_name="search",
-                content="Found result",
-            ),
-        ]
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        # Messages: assistant (with tool_calls) + tool result
-        assert len(messages) == 2
-        assert messages[0]["role"] == "assistant"
-        assert messages[0]["content"] == "Let me search..."
-        assert len(messages[0]["tool_calls"]) == 1
-        assert messages[0]["tool_calls"][0]["id"] == "tc_1"
-        assert messages[1]["role"] == "tool"
-        assert messages[1]["tool_call_id"] == "tc_1"
-        assert messages[1]["content"] == "Found result"
-
-        # Transcript: user + assistant(tool_use) + user(tool_result)
-        assert builder.entry_count == 3
-
-    def test_tool_calls_without_text(self):
-        """Tool calls without accompanying text should still work."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[
-                LLMToolCall(id="tc_1", name="run", arguments="{}"),
-            ],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-        tool_results = [
-            ToolCallResult(tool_call_id="tc_1", tool_name="run", content="done"),
-        ]
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert len(messages) == 2
-        assert "content" not in messages[0]  # No text content
-        assert messages[0]["tool_calls"][0]["function"]["name"] == "run"
-
-    def test_no_text_no_tools(self):
-        """When the response has no text and no tool calls, nothing is appended."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=None,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert len(messages) == 0
-        # Only the user entry from setup
-        assert builder.entry_count == 1
-
-    def test_multiple_tool_calls(self):
-        """Multiple tool calls in a single response are all recorded."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[
-                LLMToolCall(id="tc_1", name="tool_a", arguments="{}"),
-                LLMToolCall(id="tc_2", name="tool_b", arguments='{"x": 1}'),
-            ],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-        tool_results = [
-            ToolCallResult(tool_call_id="tc_1", tool_name="tool_a", content="result_a"),
-            ToolCallResult(tool_call_id="tc_2", tool_name="tool_b", content="result_b"),
-        ]
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        # 1 assistant + 2 tool results
-        assert len(messages) == 3
-        assert len(messages[0]["tool_calls"]) == 2
-        assert messages[1]["tool_call_id"] == "tc_1"
-        assert messages[2]["tool_call_id"] == "tc_2"
-
-    def test_invalid_tool_arguments_handled(self):
-        """Tool call with invalid JSON arguments: the arguments field is
-        stored as-is in the message, and orjson failure falls back to {}
-        in the transcript content_blocks."""
-        messages: list = []
-        builder = self._make_transcript_builder()
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[
-                LLMToolCall(id="tc_1", name="tool_x", arguments="not-json"),
-            ],
-            raw_response=None,
-            prompt_tokens=0,
-            completion_tokens=0,
-        )
-        tool_results = [
-            ToolCallResult(tool_call_id="tc_1", tool_name="tool_x", content="ok"),
-        ]
-
-        _baseline_conversation_updater(
-            messages,
-            response,
-            tool_results=tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        # Should not raise — invalid JSON falls back to {} in transcript
-        assert len(messages) == 2
-        assert messages[0]["tool_calls"][0]["function"]["arguments"] == "not-json"
-
-
-class TestCompressSessionMessagesPreservesToolCalls:
-    """``_compress_session_messages`` must round-trip tool_calls + tool_call_id.
-
-    Compression serialises ChatMessage to dict for ``compress_context`` and
-    reifies the result back to ChatMessage.  A regression that drops
-    ``tool_calls`` or ``tool_call_id`` would corrupt the OpenAI message
-    list and break downstream tool-execution rounds.
-    """
-
-    @pytest.mark.asyncio
-    async def test_compressed_output_keeps_tool_calls_and_ids(self):
-        # Simulate compression that returns a summary + the most recent
-        # assistant(tool_call) + tool(tool_result) intact.
-        summary = {"role": "system", "content": "prior turns: user asked X"}
-        assistant_with_tc = {
-            "role": "assistant",
-            "content": "calling tool",
-            "tool_calls": [
-                {
-                    "id": "tc_abc",
-                    "type": "function",
-                    "function": {"name": "search", "arguments": '{"q":"y"}'},
-                }
-            ],
-        }
-        tool_result = {
-            "role": "tool",
-            "tool_call_id": "tc_abc",
-            "content": "search result",
-        }
-
-        compress_result = CompressResult(
-            messages=[summary, assistant_with_tc, tool_result],
-            token_count=100,
-            was_compacted=True,
-            original_token_count=5000,
-            messages_summarized=10,
-            messages_dropped=0,
-        )
-
-        # Input: messages that should be compressed.
-        input_messages = [
-            ChatMessage(role="user", content="q1"),
-            ChatMessage(
-                role="assistant",
-                content="calling tool",
-                tool_calls=[
-                    {
-                        "id": "tc_abc",
-                        "type": "function",
-                        "function": {
-                            "name": "search",
-                            "arguments": '{"q":"y"}',
-                        },
-                    }
-                ],
-            ),
-            ChatMessage(
-                role="tool",
-                tool_call_id="tc_abc",
-                content="search result",
-            ),
-        ]
-
-        with patch(
-            "backend.copilot.baseline.service.compress_context",
-            new=AsyncMock(return_value=compress_result),
-        ):
-            compressed = await _compress_session_messages(
-                input_messages, model="openrouter/anthropic/claude-opus-4"
-            )
-
-        # Summary, assistant(tool_calls), tool(tool_call_id).
-        assert len(compressed) == 3
-        # Assistant message must keep its tool_calls intact.
-        assistant_msg = compressed[1]
-        assert assistant_msg.role == "assistant"
-        assert assistant_msg.tool_calls is not None
-        assert len(assistant_msg.tool_calls) == 1
-        assert assistant_msg.tool_calls[0]["id"] == "tc_abc"
-        assert assistant_msg.tool_calls[0]["function"]["name"] == "search"
-        # Tool-role message must keep tool_call_id for OpenAI linkage.
-        tool_msg = compressed[2]
-        assert tool_msg.role == "tool"
-        assert tool_msg.tool_call_id == "tc_abc"
-        assert tool_msg.content == "search result"
-
-    @pytest.mark.asyncio
-    async def test_uncompressed_passthrough_keeps_fields(self):
-        """When compression is a no-op (was_compacted=False), the original
-        messages must be returned unchanged — including tool_calls."""
-        input_messages = [
-            ChatMessage(
-                role="assistant",
-                content="c",
-                tool_calls=[
-                    {
-                        "id": "t1",
-                        "type": "function",
-                        "function": {"name": "f", "arguments": "{}"},
-                    }
-                ],
-            ),
-            ChatMessage(role="tool", tool_call_id="t1", content="ok"),
-        ]
-
-        noop_result = CompressResult(
-            messages=[],  # ignored when was_compacted=False
-            token_count=10,
-            was_compacted=False,
-        )
-
-        with patch(
-            "backend.copilot.baseline.service.compress_context",
-            new=AsyncMock(return_value=noop_result),
-        ):
-            out = await _compress_session_messages(
-                input_messages, model="openrouter/anthropic/claude-opus-4"
-            )
-
-        assert out is input_messages  # same list returned
-        assert out[0].tool_calls is not None
-        assert out[0].tool_calls[0]["id"] == "t1"
-        assert out[1].tool_call_id == "t1"
-
-
-# ---- _ThinkingStripper tests ---- #
-
-
-def test_thinking_stripper_basic_thinking_tag() -> None:
-    """<thinking>...</thinking> blocks are fully stripped."""
-    s = _ThinkingStripper()
-    assert s.process("<thinking>internal reasoning here</thinking>Hello!") == "Hello!"
-
-
-def test_thinking_stripper_internal_reasoning_tag() -> None:
-    """<internal_reasoning>...</internal_reasoning> blocks (Gemini) are stripped."""
-    s = _ThinkingStripper()
-    assert (
-        s.process("<internal_reasoning>step by step</internal_reasoning>Answer")
-        == "Answer"
-    )
-
-
-def test_thinking_stripper_split_across_chunks() -> None:
-    """Tags split across multiple chunks are handled correctly."""
-    s = _ThinkingStripper()
-    out = s.process("Hello <thin")
-    out += s.process("king>secret</thinking> world")
-    assert out == "Hello  world"
-
-
-def test_thinking_stripper_plain_text_preserved() -> None:
-    """Plain text with the word 'thinking' is not stripped."""
-    s = _ThinkingStripper()
-    assert (
-        s.process("I am thinking about this problem")
-        == "I am thinking about this problem"
-    )
-
-
-def test_thinking_stripper_multiple_blocks() -> None:
-    """Multiple reasoning blocks in one stream are all stripped."""
-    s = _ThinkingStripper()
-    result = s.process(
-        "A<thinking>x</thinking>B<internal_reasoning>y</internal_reasoning>C"
-    )
-    assert result == "ABC"
-
-
-def test_thinking_stripper_flush_discards_unclosed() -> None:
-    """Unclosed reasoning block is discarded on flush."""
-    s = _ThinkingStripper()
-    s.process("Start<thinking>never closed")
-    flushed = s.flush()
-    assert "never closed" not in flushed
-
-
-def test_thinking_stripper_empty_block() -> None:
-    """Empty reasoning blocks are handled gracefully."""
-    s = _ThinkingStripper()
-    assert s.process("Before<thinking></thinking>After") == "BeforeAfter"
-
-
-# ---- _filter_tools_by_permissions tests ---- #
-
-
-def _make_tool(name: str) -> ChatCompletionToolParam:
-    """Build a minimal OpenAI ChatCompletionToolParam."""
-    return ChatCompletionToolParam(
-        type="function",
-        function={"name": name, "parameters": {}},
-    )
-
-
-class TestFilterToolsByPermissions:
-    """Tests for _filter_tools_by_permissions."""
-
-    @patch(
-        "backend.copilot.permissions.all_known_tool_names",
-        return_value=frozenset({"run_block", "web_fetch", "bash_exec"}),
-    )
-    def test_empty_permissions_returns_all(self, _mock_names):
-        """Empty permissions (no filtering) returns every tool unchanged."""
-        from backend.copilot.baseline.service import _filter_tools_by_permissions
-        from backend.copilot.permissions import CopilotPermissions
-
-        tools = [_make_tool("run_block"), _make_tool("web_fetch")]
-        perms = CopilotPermissions()
-        result = _filter_tools_by_permissions(tools, perms)
-        assert result == tools
-
-    @patch(
-        "backend.copilot.permissions.all_known_tool_names",
-        return_value=frozenset({"run_block", "web_fetch", "bash_exec"}),
-    )
-    def test_allowlist_keeps_only_matching(self, _mock_names):
-        """Explicit allowlist (tools_exclude=False) keeps only listed tools."""
-        from backend.copilot.baseline.service import _filter_tools_by_permissions
-        from backend.copilot.permissions import CopilotPermissions
-
-        tools = [
-            _make_tool("run_block"),
-            _make_tool("web_fetch"),
-            _make_tool("bash_exec"),
-        ]
-        perms = CopilotPermissions(tools=["web_fetch"], tools_exclude=False)
-        result = _filter_tools_by_permissions(tools, perms)
-        assert len(result) == 1
-        assert result[0]["function"]["name"] == "web_fetch"
-
-    @patch(
-        "backend.copilot.permissions.all_known_tool_names",
-        return_value=frozenset({"run_block", "web_fetch", "bash_exec"}),
-    )
-    def test_blacklist_excludes_listed(self, _mock_names):
-        """Blacklist (tools_exclude=True) removes only the listed tools."""
-        from backend.copilot.baseline.service import _filter_tools_by_permissions
-        from backend.copilot.permissions import CopilotPermissions
-
-        tools = [
-            _make_tool("run_block"),
-            _make_tool("web_fetch"),
-            _make_tool("bash_exec"),
-        ]
-        perms = CopilotPermissions(tools=["bash_exec"], tools_exclude=True)
-        result = _filter_tools_by_permissions(tools, perms)
-        names = [t["function"]["name"] for t in result]
-        assert "bash_exec" not in names
-        assert "run_block" in names
-        assert "web_fetch" in names
-        assert len(result) == 2
-
-    @patch(
-        "backend.copilot.permissions.all_known_tool_names",
-        return_value=frozenset({"run_block", "web_fetch", "bash_exec"}),
-    )
-    def test_unknown_tool_name_filtered_out(self, _mock_names):
-        """A tool whose name is not in all_known_tool_names is dropped."""
-        from backend.copilot.baseline.service import _filter_tools_by_permissions
-        from backend.copilot.permissions import CopilotPermissions
-
-        tools = [_make_tool("run_block"), _make_tool("unknown_tool")]
-        perms = CopilotPermissions(tools=["run_block"], tools_exclude=False)
-        result = _filter_tools_by_permissions(tools, perms)
-        names = [t["function"]["name"] for t in result]
-        assert "unknown_tool" not in names
-        assert names == ["run_block"]
-
-
-# ---- _prepare_baseline_attachments tests ---- #
-
-
-class TestPrepareBaselineAttachments:
-    """Tests for _prepare_baseline_attachments."""
-
-    @pytest.mark.asyncio
-    async def test_empty_file_ids(self):
-        """Empty file_ids returns empty hint and blocks."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        hint, blocks = await _prepare_baseline_attachments([], "user1", "sess1", "/tmp")
-        assert hint == ""
-        assert blocks == []
-
-    @pytest.mark.asyncio
-    async def test_empty_user_id(self):
-        """Empty user_id returns empty hint and blocks."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        hint, blocks = await _prepare_baseline_attachments(
-            ["file1"], "", "sess1", "/tmp"
-        )
-        assert hint == ""
-        assert blocks == []
-
-    @pytest.mark.asyncio
-    async def test_image_file_returns_vision_blocks(self):
-        """A PNG image within size limits is returned as a base64 vision block."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        fake_info = AsyncMock()
-        fake_info.name = "photo.png"
-        fake_info.mime_type = "image/png"
-        fake_info.size_bytes = 1024
-
-        fake_manager = AsyncMock()
-        fake_manager.get_file_info = AsyncMock(return_value=fake_info)
-        fake_manager.read_file_by_id = AsyncMock(return_value=b"\x89PNG_FAKE_DATA")
-
-        with patch(
-            "backend.copilot.baseline.service.get_workspace_manager",
-            new=AsyncMock(return_value=fake_manager),
-        ):
-            hint, blocks = await _prepare_baseline_attachments(
-                ["fid1"], "user1", "sess1", "/tmp/workdir"
-            )
-
-        assert len(blocks) == 1
-        assert blocks[0]["type"] == "image"
-        assert blocks[0]["source"]["media_type"] == "image/png"
-        assert blocks[0]["source"]["type"] == "base64"
-        assert "photo.png" in hint
-        assert "embedded as image" in hint
-
-    @pytest.mark.asyncio
-    async def test_non_image_file_saved_to_working_dir(self, tmp_path):
-        """A non-image file is written to working_dir."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        fake_info = AsyncMock()
-        fake_info.name = "data.csv"
-        fake_info.mime_type = "text/csv"
-        fake_info.size_bytes = 42
-
-        fake_manager = AsyncMock()
-        fake_manager.get_file_info = AsyncMock(return_value=fake_info)
-        fake_manager.read_file_by_id = AsyncMock(return_value=b"col1,col2\na,b")
-
-        with patch(
-            "backend.copilot.baseline.service.get_workspace_manager",
-            new=AsyncMock(return_value=fake_manager),
-        ):
-            hint, blocks = await _prepare_baseline_attachments(
-                ["fid1"], "user1", "sess1", str(tmp_path)
-            )
-
-        assert blocks == []
-        assert "data.csv" in hint
-        assert "saved to" in hint
-        saved = tmp_path / "data.csv"
-        assert saved.exists()
-        assert saved.read_bytes() == b"col1,col2\na,b"
-
-    @pytest.mark.asyncio
-    async def test_file_not_found_skipped(self):
-        """When get_file_info returns None the file is silently skipped."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        fake_manager = AsyncMock()
-        fake_manager.get_file_info = AsyncMock(return_value=None)
-
-        with patch(
-            "backend.copilot.baseline.service.get_workspace_manager",
-            new=AsyncMock(return_value=fake_manager),
-        ):
-            hint, blocks = await _prepare_baseline_attachments(
-                ["missing_id"], "user1", "sess1", "/tmp"
-            )
-
-        assert hint == ""
-        assert blocks == []
-
-    @pytest.mark.asyncio
-    async def test_workspace_manager_error(self):
-        """When get_workspace_manager raises, returns empty results."""
-        from backend.copilot.baseline.service import _prepare_baseline_attachments
-
-        with patch(
-            "backend.copilot.baseline.service.get_workspace_manager",
-            new=AsyncMock(side_effect=RuntimeError("connection failed")),
-        ):
-            hint, blocks = await _prepare_baseline_attachments(
-                ["fid1"], "user1", "sess1", "/tmp"
-            )
-
-        assert hint == ""
-        assert blocks == []

autogpt_platform/backend/backend/copilot/baseline/transcript_integration_test.py

@@ -1,667 +0,0 @@
-"""Integration tests for baseline transcript flow.
-
-Exercises the real helpers in ``baseline/service.py`` that download,
-validate, load, append to, backfill, and upload the transcript.
-Storage is mocked via ``download_transcript`` / ``upload_transcript``
-patches; no network access is required.
-"""
-
-import json as stdlib_json
-from unittest.mock import AsyncMock, patch
-
-import pytest
-
-from backend.copilot.baseline.service import (
-    _load_prior_transcript,
-    _record_turn_to_transcript,
-    _resolve_baseline_model,
-    _upload_final_transcript,
-    is_transcript_stale,
-    should_upload_transcript,
-)
-from backend.copilot.service import config
-from backend.copilot.transcript import (
-    STOP_REASON_END_TURN,
-    STOP_REASON_TOOL_USE,
-    TranscriptDownload,
-)
-from backend.copilot.transcript_builder import TranscriptBuilder
-from backend.util.tool_call_loop import LLMLoopResponse, LLMToolCall, ToolCallResult
-
-
-def _make_transcript_content(*roles: str) -> str:
-    """Build a minimal valid JSONL transcript from role names."""
-    lines = []
-    parent = ""
-    for i, role in enumerate(roles):
-        uid = f"uuid-{i}"
-        entry: dict = {
-            "type": role,
-            "uuid": uid,
-            "parentUuid": parent,
-            "message": {
-                "role": role,
-                "content": [{"type": "text", "text": f"{role} message {i}"}],
-            },
-        }
-        if role == "assistant":
-            entry["message"]["id"] = f"msg_{i}"
-            entry["message"]["model"] = "test-model"
-            entry["message"]["type"] = "message"
-            entry["message"]["stop_reason"] = STOP_REASON_END_TURN
-        lines.append(stdlib_json.dumps(entry))
-        parent = uid
-    return "\n".join(lines) + "\n"
-
-
-class TestResolveBaselineModel:
-    """Model selection honours the per-request mode."""
-
-    def test_fast_mode_selects_fast_model(self):
-        assert _resolve_baseline_model("fast") == config.fast_model
-
-    def test_extended_thinking_selects_default_model(self):
-        assert _resolve_baseline_model("extended_thinking") == config.model
-
-    def test_none_mode_selects_default_model(self):
-        """Critical: baseline users without a mode MUST keep the default (opus)."""
-        assert _resolve_baseline_model(None) == config.model
-
-    def test_default_and_fast_models_differ(self):
-        """Sanity: the two tiers are actually distinct in production config."""
-        assert config.model != config.fast_model
-
-
-class TestLoadPriorTranscript:
-    """``_load_prior_transcript`` wraps the download + validate + load flow."""
-
-    @pytest.mark.asyncio
-    async def test_loads_fresh_transcript(self):
-        builder = TranscriptBuilder()
-        content = _make_transcript_content("user", "assistant")
-        download = TranscriptDownload(content=content, message_count=2)
-
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=download),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=3,
-                transcript_builder=builder,
-            )
-
-        assert covers is True
-        assert builder.entry_count == 2
-        assert builder.last_entry_type == "assistant"
-
-    @pytest.mark.asyncio
-    async def test_rejects_stale_transcript(self):
-        """msg_count strictly less than session-1 is treated as stale."""
-        builder = TranscriptBuilder()
-        content = _make_transcript_content("user", "assistant")
-        # session has 6 messages, transcript only covers 2 → stale.
-        download = TranscriptDownload(content=content, message_count=2)
-
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=download),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=6,
-                transcript_builder=builder,
-            )
-
-        assert covers is False
-        assert builder.is_empty
-
-    @pytest.mark.asyncio
-    async def test_missing_transcript_returns_false(self):
-        builder = TranscriptBuilder()
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=None),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=2,
-                transcript_builder=builder,
-            )
-
-        assert covers is False
-        assert builder.is_empty
-
-    @pytest.mark.asyncio
-    async def test_invalid_transcript_returns_false(self):
-        builder = TranscriptBuilder()
-        download = TranscriptDownload(
-            content='{"type":"progress","uuid":"a"}\n',
-            message_count=1,
-        )
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=download),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=2,
-                transcript_builder=builder,
-            )
-
-        assert covers is False
-        assert builder.is_empty
-
-    @pytest.mark.asyncio
-    async def test_download_exception_returns_false(self):
-        builder = TranscriptBuilder()
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(side_effect=RuntimeError("boom")),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=2,
-                transcript_builder=builder,
-            )
-
-        assert covers is False
-        assert builder.is_empty
-
-    @pytest.mark.asyncio
-    async def test_zero_message_count_not_stale(self):
-        """When msg_count is 0 (unknown), staleness check is skipped."""
-        builder = TranscriptBuilder()
-        download = TranscriptDownload(
-            content=_make_transcript_content("user", "assistant"),
-            message_count=0,
-        )
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=download),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=20,
-                transcript_builder=builder,
-            )
-
-        assert covers is True
-        assert builder.entry_count == 2
-
-
-class TestUploadFinalTranscript:
-    """``_upload_final_transcript`` serialises and calls storage."""
-
-    @pytest.mark.asyncio
-    async def test_uploads_valid_transcript(self):
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-        builder.append_assistant(
-            content_blocks=[{"type": "text", "text": "hello"}],
-            model="test-model",
-            stop_reason=STOP_REASON_END_TURN,
-        )
-
-        upload_mock = AsyncMock(return_value=None)
-        with patch(
-            "backend.copilot.baseline.service.upload_transcript",
-            new=upload_mock,
-        ):
-            await _upload_final_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                transcript_builder=builder,
-                session_msg_count=2,
-            )
-
-        upload_mock.assert_awaited_once()
-        assert upload_mock.await_args is not None
-        call_kwargs = upload_mock.await_args.kwargs
-        assert call_kwargs["user_id"] == "user-1"
-        assert call_kwargs["session_id"] == "session-1"
-        assert call_kwargs["message_count"] == 2
-        assert "hello" in call_kwargs["content"]
-
-    @pytest.mark.asyncio
-    async def test_skips_upload_when_builder_empty(self):
-        builder = TranscriptBuilder()
-        upload_mock = AsyncMock(return_value=None)
-        with patch(
-            "backend.copilot.baseline.service.upload_transcript",
-            new=upload_mock,
-        ):
-            await _upload_final_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                transcript_builder=builder,
-                session_msg_count=0,
-            )
-
-        upload_mock.assert_not_awaited()
-
-    @pytest.mark.asyncio
-    async def test_swallows_upload_exceptions(self):
-        """Upload failures should not propagate (flow continues for the user)."""
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-        builder.append_assistant(
-            content_blocks=[{"type": "text", "text": "hello"}],
-            model="test-model",
-            stop_reason=STOP_REASON_END_TURN,
-        )
-
-        with patch(
-            "backend.copilot.baseline.service.upload_transcript",
-            new=AsyncMock(side_effect=RuntimeError("storage unavailable")),
-        ):
-            # Should not raise.
-            await _upload_final_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                transcript_builder=builder,
-                session_msg_count=2,
-            )
-
-
-class TestRecordTurnToTranscript:
-    """``_record_turn_to_transcript`` translates LLMLoopResponse → transcript."""
-
-    def test_records_final_assistant_text(self):
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-
-        response = LLMLoopResponse(
-            response_text="hello there",
-            tool_calls=[],
-            raw_response=None,
-        )
-        _record_turn_to_transcript(
-            response,
-            tool_results=None,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert builder.entry_count == 2
-        assert builder.last_entry_type == "assistant"
-        jsonl = builder.to_jsonl()
-        assert "hello there" in jsonl
-        assert STOP_REASON_END_TURN in jsonl
-
-    def test_records_tool_use_then_tool_result(self):
-        """Anthropic ordering: assistant(tool_use) → user(tool_result)."""
-        builder = TranscriptBuilder()
-        builder.append_user(content="use a tool")
-
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[
-                LLMToolCall(id="call-1", name="echo", arguments='{"text":"hi"}')
-            ],
-            raw_response=None,
-        )
-        tool_results = [
-            ToolCallResult(tool_call_id="call-1", tool_name="echo", content="hi")
-        ]
-        _record_turn_to_transcript(
-            response,
-            tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        # user, assistant(tool_use), user(tool_result) = 3 entries
-        assert builder.entry_count == 3
-        jsonl = builder.to_jsonl()
-        assert STOP_REASON_TOOL_USE in jsonl
-        assert "tool_use" in jsonl
-        assert "tool_result" in jsonl
-        assert "call-1" in jsonl
-
-    def test_records_nothing_on_empty_response(self):
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[],
-            raw_response=None,
-        )
-        _record_turn_to_transcript(
-            response,
-            tool_results=None,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert builder.entry_count == 1
-
-    def test_malformed_tool_args_dont_crash(self):
-        """Bad JSON in tool arguments falls back to {} without raising."""
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-
-        response = LLMLoopResponse(
-            response_text=None,
-            tool_calls=[LLMToolCall(id="call-1", name="echo", arguments="{not-json")],
-            raw_response=None,
-        )
-        tool_results = [
-            ToolCallResult(tool_call_id="call-1", tool_name="echo", content="ok")
-        ]
-        _record_turn_to_transcript(
-            response,
-            tool_results,
-            transcript_builder=builder,
-            model="test-model",
-        )
-
-        assert builder.entry_count == 3
-        jsonl = builder.to_jsonl()
-        assert '"input":{}' in jsonl
-
-
-class TestRoundTrip:
-    """End-to-end: load prior → append new turn → upload."""
-
-    @pytest.mark.asyncio
-    async def test_full_round_trip(self):
-        prior = _make_transcript_content("user", "assistant")
-        download = TranscriptDownload(content=prior, message_count=2)
-
-        builder = TranscriptBuilder()
-        with patch(
-            "backend.copilot.baseline.service.download_transcript",
-            new=AsyncMock(return_value=download),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=3,
-                transcript_builder=builder,
-            )
-        assert covers is True
-        assert builder.entry_count == 2
-
-        # New user turn.
-        builder.append_user(content="new question")
-        assert builder.entry_count == 3
-
-        # New assistant turn.
-        response = LLMLoopResponse(
-            response_text="new answer",
-            tool_calls=[],
-            raw_response=None,
-        )
-        _record_turn_to_transcript(
-            response,
-            tool_results=None,
-            transcript_builder=builder,
-            model="test-model",
-        )
-        assert builder.entry_count == 4
-
-        # Upload.
-        upload_mock = AsyncMock(return_value=None)
-        with patch(
-            "backend.copilot.baseline.service.upload_transcript",
-            new=upload_mock,
-        ):
-            await _upload_final_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                transcript_builder=builder,
-                session_msg_count=4,
-            )
-
-        upload_mock.assert_awaited_once()
-        assert upload_mock.await_args is not None
-        uploaded = upload_mock.await_args.kwargs["content"]
-        assert "new question" in uploaded
-        assert "new answer" in uploaded
-        # Original content preserved in the round trip.
-        assert "user message 0" in uploaded
-        assert "assistant message 1" in uploaded
-
-    @pytest.mark.asyncio
-    async def test_backfill_append_guard(self):
-        """Backfill only runs when the last entry is not already assistant."""
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-
-        # Simulate the backfill guard from stream_chat_completion_baseline.
-        assistant_text = "partial text before error"
-        if builder.last_entry_type != "assistant":
-            builder.append_assistant(
-                content_blocks=[{"type": "text", "text": assistant_text}],
-                model="test-model",
-                stop_reason=STOP_REASON_END_TURN,
-            )
-
-        assert builder.last_entry_type == "assistant"
-        assert "partial text before error" in builder.to_jsonl()
-
-        # Second invocation: the guard must prevent double-append.
-        initial_count = builder.entry_count
-        if builder.last_entry_type != "assistant":
-            builder.append_assistant(
-                content_blocks=[{"type": "text", "text": "duplicate"}],
-                model="test-model",
-                stop_reason=STOP_REASON_END_TURN,
-            )
-        assert builder.entry_count == initial_count
-
-
-class TestIsTranscriptStale:
-    """``is_transcript_stale`` gates prior-transcript loading."""
-
-    def test_none_download_is_not_stale(self):
-        assert is_transcript_stale(None, session_msg_count=5) is False
-
-    def test_zero_message_count_is_not_stale(self):
-        """Legacy transcripts without msg_count tracking must remain usable."""
-        dl = TranscriptDownload(content="", message_count=0)
-        assert is_transcript_stale(dl, session_msg_count=20) is False
-
-    def test_stale_when_covers_less_than_prefix(self):
-        dl = TranscriptDownload(content="", message_count=2)
-        # session has 6 messages; transcript must cover at least 5 (6-1).
-        assert is_transcript_stale(dl, session_msg_count=6) is True
-
-    def test_fresh_when_covers_full_prefix(self):
-        dl = TranscriptDownload(content="", message_count=5)
-        assert is_transcript_stale(dl, session_msg_count=6) is False
-
-    def test_fresh_when_exceeds_prefix(self):
-        """Race: transcript ahead of session count is still acceptable."""
-        dl = TranscriptDownload(content="", message_count=10)
-        assert is_transcript_stale(dl, session_msg_count=6) is False
-
-    def test_boundary_equal_to_prefix_minus_one(self):
-        dl = TranscriptDownload(content="", message_count=5)
-        assert is_transcript_stale(dl, session_msg_count=6) is False
-
-
-class TestShouldUploadTranscript:
-    """``should_upload_transcript`` gates the final upload."""
-
-    def test_upload_allowed_for_user_with_coverage(self):
-        assert should_upload_transcript("user-1", True) is True
-
-    def test_upload_skipped_when_no_user(self):
-        assert should_upload_transcript(None, True) is False
-
-    def test_upload_skipped_when_empty_user(self):
-        assert should_upload_transcript("", True) is False
-
-    def test_upload_skipped_without_coverage(self):
-        """Partial transcript must never clobber a more complete stored one."""
-        assert should_upload_transcript("user-1", False) is False
-
-    def test_upload_skipped_when_no_user_and_no_coverage(self):
-        assert should_upload_transcript(None, False) is False
-
-
-class TestTranscriptLifecycle:
-    """End-to-end: download → validate → build → upload.
-
-    Simulates the full transcript lifecycle inside
-    ``stream_chat_completion_baseline`` by mocking the storage layer and
-    driving each step through the real helpers.
-    """
-
-    @pytest.mark.asyncio
-    async def test_full_lifecycle_happy_path(self):
-        """Fresh download, append a turn, upload covers the session."""
-        builder = TranscriptBuilder()
-        prior = _make_transcript_content("user", "assistant")
-        download = TranscriptDownload(content=prior, message_count=2)
-
-        upload_mock = AsyncMock(return_value=None)
-        with (
-            patch(
-                "backend.copilot.baseline.service.download_transcript",
-                new=AsyncMock(return_value=download),
-            ),
-            patch(
-                "backend.copilot.baseline.service.upload_transcript",
-                new=upload_mock,
-            ),
-        ):
-            # --- 1. Download & load prior transcript ---
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=3,
-                transcript_builder=builder,
-            )
-            assert covers is True
-
-            # --- 2. Append a new user turn + a new assistant response ---
-            builder.append_user(content="follow-up question")
-            _record_turn_to_transcript(
-                LLMLoopResponse(
-                    response_text="follow-up answer",
-                    tool_calls=[],
-                    raw_response=None,
-                ),
-                tool_results=None,
-                transcript_builder=builder,
-                model="test-model",
-            )
-
-            # --- 3. Gate + upload ---
-            assert (
-                should_upload_transcript(
-                    user_id="user-1", transcript_covers_prefix=covers
-                )
-                is True
-            )
-            await _upload_final_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                transcript_builder=builder,
-                session_msg_count=4,
-            )
-
-        upload_mock.assert_awaited_once()
-        assert upload_mock.await_args is not None
-        uploaded = upload_mock.await_args.kwargs["content"]
-        assert "follow-up question" in uploaded
-        assert "follow-up answer" in uploaded
-        # Original prior-turn content preserved.
-        assert "user message 0" in uploaded
-        assert "assistant message 1" in uploaded
-
-    @pytest.mark.asyncio
-    async def test_lifecycle_stale_download_suppresses_upload(self):
-        """Stale download → covers=False → upload must be skipped."""
-        builder = TranscriptBuilder()
-        # session has 10 msgs but stored transcript only covers 2 → stale.
-        stale = TranscriptDownload(
-            content=_make_transcript_content("user", "assistant"),
-            message_count=2,
-        )
-
-        upload_mock = AsyncMock(return_value=None)
-        with (
-            patch(
-                "backend.copilot.baseline.service.download_transcript",
-                new=AsyncMock(return_value=stale),
-            ),
-            patch(
-                "backend.copilot.baseline.service.upload_transcript",
-                new=upload_mock,
-            ),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=10,
-                transcript_builder=builder,
-            )
-
-        assert covers is False
-        # The caller's gate mirrors the production path.
-        assert (
-            should_upload_transcript(user_id="user-1", transcript_covers_prefix=covers)
-            is False
-        )
-        upload_mock.assert_not_awaited()
-
-    @pytest.mark.asyncio
-    async def test_lifecycle_anonymous_user_skips_upload(self):
-        """Anonymous (user_id=None) → upload gate must return False."""
-        builder = TranscriptBuilder()
-        builder.append_user(content="hi")
-        builder.append_assistant(
-            content_blocks=[{"type": "text", "text": "hello"}],
-            model="test-model",
-            stop_reason=STOP_REASON_END_TURN,
-        )
-
-        assert (
-            should_upload_transcript(user_id=None, transcript_covers_prefix=True)
-            is False
-        )
-
-    @pytest.mark.asyncio
-    async def test_lifecycle_missing_download_still_uploads_new_content(self):
-        """No prior transcript → covers defaults to True in the service,
-        new turn should upload cleanly."""
-        builder = TranscriptBuilder()
-        upload_mock = AsyncMock(return_value=None)
-        with (
-            patch(
-                "backend.copilot.baseline.service.download_transcript",
-                new=AsyncMock(return_value=None),
-            ),
-            patch(
-                "backend.copilot.baseline.service.upload_transcript",
-                new=upload_mock,
-            ),
-        ):
-            covers = await _load_prior_transcript(
-                user_id="user-1",
-                session_id="session-1",
-                session_msg_count=1,
-                transcript_builder=builder,
-            )
-            # No download: covers is False, so the production path would
-            # skip upload. This protects against overwriting a future
-            # more-complete transcript with a single-turn snapshot.
-            assert covers is False
-            assert (
-                should_upload_transcript(
-                    user_id="user-1", transcript_covers_prefix=covers
-                )
-                is False
-            )
-            upload_mock.assert_not_awaited()

autogpt_platform/backend/backend/copilot/config.py

@@ -8,14 +8,6 @@ from pydantic_settings import BaseSettings
 
 from backend.util.clients import OPENROUTER_BASE_URL
 
-# Per-request routing mode for a single chat turn.
-# - 'fast': route to the baseline OpenAI-compatible path with the cheaper model.
-# - 'extended_thinking': route to the Claude Agent SDK path with the default
-#   (opus) model.
-# ``None`` means "no override"; the server falls back to the Claude Code
-# subscription flag → LaunchDarkly COPILOT_SDK → config.use_claude_agent_sdk.
-CopilotMode = Literal["fast", "extended_thinking"]
-
 
 class ChatConfig(BaseSettings):
     """Configuration for the chat system."""
@@ -146,6 +138,32 @@ class ChatConfig(BaseSettings):
         description="Use --resume for multi-turn conversations instead of "
         "history compression. Falls back to compression when unavailable.",
     )
+    claude_agent_fallback_model: str = Field(
+        default="claude-sonnet-4-20250514",
+        description="Fallback model when the primary model is unavailable (e.g. 529 "
+        "overloaded). The SDK automatically retries with this cheaper model.",
+    )
+    claude_agent_max_turns: int = Field(
+        default=50,
+        ge=1,
+        le=500,
+        description="Maximum number of agentic turns (tool-use loops) per query. "
+        "Prevents runaway tool loops from burning budget.",
+    )
+    claude_agent_max_budget_usd: float = Field(
+        default=5.0,
+        ge=0.01,
+        le=100.0,
+        description="Maximum spend in USD per SDK query. The CLI aborts the "
+        "request if this budget is exceeded.",
+    )
+    claude_agent_max_transient_retries: int = Field(
+        default=3,
+        ge=0,
+        le=10,
+        description="Maximum number of retries for transient API errors "
+        "(429, 5xx, ECONNRESET) before surfacing the error to the user.",
+    )
     use_openrouter: bool = Field(
         default=True,
         description="Enable routing API calls through the OpenRouter proxy. "

autogpt_platform/backend/backend/copilot/constants.py

@@ -44,12 +44,32 @@ def parse_node_id_from_exec_id(node_exec_id: str) -> str:
 # Transient Anthropic API error detection
 # ---------------------------------------------------------------------------
 # Patterns in error text that indicate a transient Anthropic API error
-# (ECONNRESET / dropped TCP connection) which is retryable.
+# which is retryable.  Covers:
+#   - Connection-level: ECONNRESET, dropped TCP connections
+#   - HTTP 429: rate-limit / too-many-requests
+#   - HTTP 5xx: server errors, overloaded
 _TRANSIENT_ERROR_PATTERNS = (
+    # Connection-level
     "socket connection was closed unexpectedly",
     "ECONNRESET",
     "connection was forcibly closed",
     "network socket disconnected",
+    # 429 rate-limit patterns
+    "rate limit",
+    "rate_limit",
+    "too many requests",
+    "status code 429",
+    # 5xx server error patterns
+    "overloaded",
+    "internal server error",
+    "bad gateway",
+    "service unavailable",
+    "gateway timeout",
+    "status code 529",
+    "status code 500",
+    "status code 502",
+    "status code 503",
+    "status code 504",
 )
 
 FRIENDLY_TRANSIENT_MSG = "Anthropic connection interrupted — please retry"

autogpt_platform/backend/backend/copilot/db.py

@@ -14,7 +14,6 @@ from prisma.types import (
     ChatSessionUpdateInput,
     ChatSessionWhereInput,
 )
-from pydantic import BaseModel
 
 from backend.data import db
 from backend.util.json import SafeJson, sanitize_string
@@ -24,22 +23,12 @@ from .model import (
     ChatSession,
     ChatSessionInfo,
     ChatSessionMetadata,
-    cache_chat_session,
+    invalidate_session_cache,
 )
-from .model import get_chat_session as get_chat_session_cached
 
 logger = logging.getLogger(__name__)
 
 
-class PaginatedMessages(BaseModel):
-    """Result of a paginated message query."""
-
-    messages: list[ChatMessage]
-    has_more: bool
-    oldest_sequence: int | None
-    session: ChatSessionInfo
-
-
 async def get_chat_session(session_id: str) -> ChatSession | None:
     """Get a chat session by ID from the database."""
     session = await PrismaChatSession.prisma().find_unique(
@@ -49,116 +38,6 @@ async def get_chat_session(session_id: str) -> ChatSession | None:
     return ChatSession.from_db(session) if session else None
 
 
-async def get_chat_session_metadata(session_id: str) -> ChatSessionInfo | None:
-    """Get chat session metadata (without messages) for ownership validation."""
-    session = await PrismaChatSession.prisma().find_unique(
-        where={"id": session_id},
-    )
-    return ChatSessionInfo.from_db(session) if session else None
-
-
-async def get_chat_messages_paginated(
-    session_id: str,
-    limit: int = 50,
-    before_sequence: int | None = None,
-    user_id: str | None = None,
-) -> PaginatedMessages | None:
-    """Get paginated messages for a session, newest first.
-
-    Verifies session existence (and ownership when ``user_id`` is provided)
-    in parallel with the message query.  Returns ``None`` when the session
-    is not found or does not belong to the user.
-
-    Args:
-        session_id: The chat session ID.
-        limit: Max messages to return.
-        before_sequence: Cursor — return messages with sequence < this value.
-        user_id: If provided, filters via ``Session.userId`` so only the
-            session owner's messages are returned (acts as an ownership guard).
-    """
-    # Build session-existence / ownership check
-    session_where: ChatSessionWhereInput = {"id": session_id}
-    if user_id is not None:
-        session_where["userId"] = user_id
-
-    # Build message include — fetch paginated messages in the same query
-    msg_include: dict[str, Any] = {
-        "order_by": {"sequence": "desc"},
-        "take": limit + 1,
-    }
-    if before_sequence is not None:
-        msg_include["where"] = {"sequence": {"lt": before_sequence}}
-
-    # Single query: session existence/ownership + paginated messages
-    session = await PrismaChatSession.prisma().find_first(
-        where=session_where,
-        include={"Messages": msg_include},
-    )
-
-    if session is None:
-        return None
-
-    session_info = ChatSessionInfo.from_db(session)
-    results = list(session.Messages) if session.Messages else []
-
-    has_more = len(results) > limit
-    results = results[:limit]
-
-    # Reverse to ascending order
-    results.reverse()
-
-    # Tool-call boundary fix: if the oldest message is a tool message,
-    # expand backward to include the preceding assistant message that
-    # owns the tool_calls, so convertChatSessionMessagesToUiMessages
-    # can pair them correctly.
-    _BOUNDARY_SCAN_LIMIT = 10
-    if results and results[0].role == "tool":
-        boundary_where: dict[str, Any] = {
-            "sessionId": session_id,
-            "sequence": {"lt": results[0].sequence},
-        }
-        if user_id is not None:
-            boundary_where["Session"] = {"is": {"userId": user_id}}
-        extra = await PrismaChatMessage.prisma().find_many(
-            where=boundary_where,
-            order={"sequence": "desc"},
-            take=_BOUNDARY_SCAN_LIMIT,
-        )
-        # Find the first non-tool message (should be the assistant)
-        boundary_msgs = []
-        found_owner = False
-        for msg in extra:
-            boundary_msgs.append(msg)
-            if msg.role != "tool":
-                found_owner = True
-                break
-        boundary_msgs.reverse()
-        if not found_owner:
-            logger.warning(
-                "Boundary expansion did not find owning assistant message "
-                "for session=%s before sequence=%s (%d msgs scanned)",
-                session_id,
-                results[0].sequence,
-                len(extra),
-            )
-        if boundary_msgs:
-            results = boundary_msgs + results
-            # Only mark has_more if the expanded boundary isn't the
-            # very start of the conversation (sequence 0).
-            if boundary_msgs[0].sequence > 0:
-                has_more = True
-
-    messages = [ChatMessage.from_db(m) for m in results]
-    oldest_sequence = messages[0].sequence if messages else None
-
-    return PaginatedMessages(
-        messages=messages,
-        has_more=has_more,
-        oldest_sequence=oldest_sequence,
-        session=session_info,
-    )
-
-
 async def create_chat_session(
     session_id: str,
     user_id: str,
@@ -501,11 +380,8 @@ async def update_tool_message_content(
 async def set_turn_duration(session_id: str, duration_ms: int) -> None:
     """Set durationMs on the last assistant message in a session.
 
-    Updates the Redis cache in-place instead of invalidating it.
-    Invalidation would delete the key, creating a window where concurrent
-    ``get_chat_session`` calls re-populate the cache from DB — potentially
-    with stale data if the DB write from the previous turn hasn't propagated.
-    This race caused duplicate user messages on the next turn.
+    Also invalidates the Redis session cache so the next GET returns
+    the updated duration.
     """
     last_msg = await PrismaChatMessage.prisma().find_first(
         where={"sessionId": session_id, "role": "assistant"},
@@ -516,13 +392,5 @@ async def set_turn_duration(session_id: str, duration_ms: int) -> None:
             where={"id": last_msg.id},
             data={"durationMs": duration_ms},
         )
-        # Update cache in-place rather than invalidating to avoid a
-        # race window where the empty cache gets re-populated with
-        # stale data by a concurrent get_chat_session call.
-        session = await get_chat_session_cached(session_id)
-        if session and session.messages:
-            for msg in reversed(session.messages):
-                if msg.role == "assistant":
-                    msg.duration_ms = duration_ms
-                    break
-            await cache_chat_session(session)
+        # Invalidate cache so the session is re-fetched from DB with durationMs
+        await invalidate_session_cache(session_id)

autogpt_platform/backend/backend/copilot/db_test.py

@@ -1,388 +0,0 @@
-"""Unit tests for copilot.db — paginated message queries."""
-
-from __future__ import annotations
-
-from datetime import UTC, datetime
-from typing import Any
-from unittest.mock import AsyncMock, patch
-
-import pytest
-from prisma.models import ChatMessage as PrismaChatMessage
-from prisma.models import ChatSession as PrismaChatSession
-
-from backend.copilot.db import (
-    PaginatedMessages,
-    get_chat_messages_paginated,
-    set_turn_duration,
-)
-from backend.copilot.model import ChatMessage as CopilotChatMessage
-from backend.copilot.model import ChatSession, get_chat_session, upsert_chat_session
-
-
-def _make_msg(
-    sequence: int,
-    role: str = "assistant",
-    content: str | None = "hello",
-    tool_calls: Any = None,
-) -> PrismaChatMessage:
-    """Build a minimal PrismaChatMessage for testing."""
-    return PrismaChatMessage(
-        id=f"msg-{sequence}",
-        createdAt=datetime.now(UTC),
-        sessionId="sess-1",
-        role=role,
-        content=content,
-        sequence=sequence,
-        toolCalls=tool_calls,
-        name=None,
-        toolCallId=None,
-        refusal=None,
-        functionCall=None,
-    )
-
-
-def _make_session(
-    session_id: str = "sess-1",
-    user_id: str = "user-1",
-    messages: list[PrismaChatMessage] | None = None,
-) -> PrismaChatSession:
-    """Build a minimal PrismaChatSession for testing."""
-    now = datetime.now(UTC)
-    session = PrismaChatSession.model_construct(
-        id=session_id,
-        createdAt=now,
-        updatedAt=now,
-        userId=user_id,
-        credentials={},
-        successfulAgentRuns={},
-        successfulAgentSchedules={},
-        totalPromptTokens=0,
-        totalCompletionTokens=0,
-        title=None,
-        metadata={},
-        Messages=messages or [],
-    )
-    return session
-
-
-SESSION_ID = "sess-1"
-
-
-@pytest.fixture()
-def mock_db():
-    """Patch ChatSession.prisma().find_first and ChatMessage.prisma().find_many.
-
-    find_first is used for the main query (session + included messages).
-    find_many is used only for boundary expansion queries.
-    """
-    with (
-        patch.object(PrismaChatSession, "prisma") as mock_session_prisma,
-        patch.object(PrismaChatMessage, "prisma") as mock_msg_prisma,
-    ):
-        find_first = AsyncMock()
-        mock_session_prisma.return_value.find_first = find_first
-
-        find_many = AsyncMock(return_value=[])
-        mock_msg_prisma.return_value.find_many = find_many
-
-        yield find_first, find_many
-
-
-# ---------- Basic pagination ----------
-
-
-@pytest.mark.asyncio
-async def test_basic_page_returns_messages_ascending(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """Messages are returned in ascending sequence order."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(3), _make_msg(2), _make_msg(1)],
-    )
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert isinstance(page, PaginatedMessages)
-    assert [m.sequence for m in page.messages] == [1, 2, 3]
-    assert page.has_more is False
-    assert page.oldest_sequence == 1
-
-
-@pytest.mark.asyncio
-async def test_has_more_when_results_exceed_limit(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """has_more is True when DB returns more than limit items."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(3), _make_msg(2), _make_msg(1)],
-    )
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=2)
-
-    assert page is not None
-    assert page.has_more is True
-    assert len(page.messages) == 2
-    assert [m.sequence for m in page.messages] == [2, 3]
-
-
-@pytest.mark.asyncio
-async def test_empty_session_returns_no_messages(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(messages=[])
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=50)
-
-    assert page is not None
-    assert page.messages == []
-    assert page.has_more is False
-    assert page.oldest_sequence is None
-
-
-@pytest.mark.asyncio
-async def test_before_sequence_filters_correctly(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """before_sequence is passed as a where filter inside the Messages include."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(2), _make_msg(1)],
-    )
-
-    await get_chat_messages_paginated(SESSION_ID, limit=50, before_sequence=5)
-
-    call_kwargs = find_first.call_args
-    include = call_kwargs.kwargs.get("include") or call_kwargs[1].get("include")
-    assert include["Messages"]["where"] == {"sequence": {"lt": 5}}
-
-
-@pytest.mark.asyncio
-async def test_no_where_on_messages_without_before_sequence(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """Without before_sequence, the Messages include has no where clause."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(messages=[_make_msg(1)])
-
-    await get_chat_messages_paginated(SESSION_ID, limit=50)
-
-    call_kwargs = find_first.call_args
-    include = call_kwargs.kwargs.get("include") or call_kwargs[1].get("include")
-    assert "where" not in include["Messages"]
-
-
-@pytest.mark.asyncio
-async def test_user_id_filter_applied_to_session_where(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """user_id adds a userId filter to the session-level where clause."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(messages=[_make_msg(1)])
-
-    await get_chat_messages_paginated(SESSION_ID, limit=50, user_id="user-abc")
-
-    call_kwargs = find_first.call_args
-    where = call_kwargs.kwargs.get("where") or call_kwargs[1].get("where")
-    assert where["userId"] == "user-abc"
-
-
-@pytest.mark.asyncio
-async def test_session_not_found_returns_none(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """Returns None when session doesn't exist or user doesn't own it."""
-    find_first, _ = mock_db
-    find_first.return_value = None
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=50)
-
-    assert page is None
-
-
-@pytest.mark.asyncio
-async def test_session_info_included_in_result(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """PaginatedMessages includes session metadata."""
-    find_first, _ = mock_db
-    find_first.return_value = _make_session(messages=[_make_msg(1)])
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=50)
-
-    assert page is not None
-    assert page.session.session_id == SESSION_ID
-
-
-# ---------- Backward boundary expansion ----------
-
-
-@pytest.mark.asyncio
-async def test_boundary_expansion_includes_assistant(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """When page starts with a tool message, expand backward to include
-    the owning assistant message."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(5, role="tool"), _make_msg(4, role="tool")],
-    )
-    find_many.return_value = [_make_msg(3, role="assistant")]
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert page is not None
-    assert [m.sequence for m in page.messages] == [3, 4, 5]
-    assert page.messages[0].role == "assistant"
-    assert page.oldest_sequence == 3
-
-
-@pytest.mark.asyncio
-async def test_boundary_expansion_includes_multiple_tool_msgs(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """Boundary expansion scans past consecutive tool messages to find
-    the owning assistant."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(7, role="tool")],
-    )
-    find_many.return_value = [
-        _make_msg(6, role="tool"),
-        _make_msg(5, role="tool"),
-        _make_msg(4, role="assistant"),
-    ]
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert page is not None
-    assert [m.sequence for m in page.messages] == [4, 5, 6, 7]
-    assert page.messages[0].role == "assistant"
-
-
-@pytest.mark.asyncio
-async def test_boundary_expansion_sets_has_more_when_not_at_start(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """After boundary expansion, has_more=True if expanded msgs aren't at seq 0."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(3, role="tool")],
-    )
-    find_many.return_value = [_make_msg(2, role="assistant")]
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert page is not None
-    assert page.has_more is True
-
-
-@pytest.mark.asyncio
-async def test_boundary_expansion_no_has_more_at_conversation_start(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """has_more stays False when boundary expansion reaches seq 0."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(1, role="tool")],
-    )
-    find_many.return_value = [_make_msg(0, role="assistant")]
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert page is not None
-    assert page.has_more is False
-    assert page.oldest_sequence == 0
-
-
-@pytest.mark.asyncio
-async def test_no_boundary_expansion_when_first_msg_not_tool(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """No boundary expansion when the first message is not a tool message."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(3, role="user"), _make_msg(2, role="assistant")],
-    )
-
-    page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-
-    assert page is not None
-    assert find_many.call_count == 0
-    assert [m.sequence for m in page.messages] == [2, 3]
-
-
-@pytest.mark.asyncio
-async def test_boundary_expansion_warns_when_no_owner_found(
-    mock_db: tuple[AsyncMock, AsyncMock],
-):
-    """When boundary scan doesn't find a non-tool message, a warning is logged
-    and the boundary messages are still included."""
-    find_first, find_many = mock_db
-    find_first.return_value = _make_session(
-        messages=[_make_msg(10, role="tool")],
-    )
-    find_many.return_value = [_make_msg(i, role="tool") for i in range(9, -1, -1)]
-
-    with patch("backend.copilot.db.logger") as mock_logger:
-        page = await get_chat_messages_paginated(SESSION_ID, limit=5)
-        mock_logger.warning.assert_called_once()
-
-    assert page is not None
-    assert page.messages[0].role == "tool"
-    assert len(page.messages) > 1
-
-
-# ---------- Turn duration (integration tests) ----------
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_set_turn_duration_updates_cache_in_place(setup_test_user, test_user_id):
-    """set_turn_duration patches the cached session without invalidation.
-
-    Verifies that after calling set_turn_duration the Redis-cached session
-    reflects the updated durationMs on the last assistant message, without
-    the cache having been deleted and re-populated (which could race with
-    concurrent get_chat_session calls).
-    """
-    session = ChatSession.new(user_id=test_user_id, dry_run=False)
-    session.messages = [
-        CopilotChatMessage(role="user", content="hello"),
-        CopilotChatMessage(role="assistant", content="hi there"),
-    ]
-    session = await upsert_chat_session(session)
-
-    # Ensure the session is in cache
-    cached = await get_chat_session(session.session_id, test_user_id)
-    assert cached is not None
-    assert cached.messages[-1].duration_ms is None
-
-    # Update turn duration — should patch cache in-place
-    await set_turn_duration(session.session_id, 1234)
-
-    # Read from cache (not DB) — the cache should already have the update
-    updated = await get_chat_session(session.session_id, test_user_id)
-    assert updated is not None
-    assistant_msgs = [m for m in updated.messages if m.role == "assistant"]
-    assert len(assistant_msgs) == 1
-    assert assistant_msgs[0].duration_ms == 1234
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_set_turn_duration_no_assistant_message(setup_test_user, test_user_id):
-    """set_turn_duration is a no-op when there are no assistant messages."""
-    session = ChatSession.new(user_id=test_user_id, dry_run=False)
-    session.messages = [
-        CopilotChatMessage(role="user", content="hello"),
-    ]
-    session = await upsert_chat_session(session)
-
-    # Should not raise
-    await set_turn_duration(session.session_id, 5678)
-
-    cached = await get_chat_session(session.session_id, test_user_id)
-    assert cached is not None
-    # User message should not have durationMs
-    assert cached.messages[0].duration_ms is None

autogpt_platform/backend/backend/copilot/executor/processor.py

@@ -13,7 +13,7 @@ import time
 
 from backend.copilot import stream_registry
 from backend.copilot.baseline import stream_chat_completion_baseline
-from backend.copilot.config import ChatConfig, CopilotMode
+from backend.copilot.config import ChatConfig
 from backend.copilot.response_model import StreamError
 from backend.copilot.sdk import service as sdk_service
 from backend.copilot.sdk.dummy import stream_chat_completion_dummy
@@ -30,57 +30,6 @@ from .utils import CoPilotExecutionEntry, CoPilotLogMetadata
 logger = TruncatedLogger(logging.getLogger(__name__), prefix="[CoPilotExecutor]")
 
 
-# ============ Mode Routing ============ #
-
-
-async def resolve_effective_mode(
-    mode: CopilotMode | None,
-    user_id: str | None,
-) -> CopilotMode | None:
-    """Strip ``mode`` when the user is not entitled to the toggle.
-
-    The UI gates the mode toggle behind ``CHAT_MODE_OPTION``; the
-    processor enforces the same gate server-side so an authenticated
-    user cannot bypass the flag by crafting a request directly.
-    """
-    if mode is None:
-        return None
-    allowed = await is_feature_enabled(
-        Flag.CHAT_MODE_OPTION,
-        user_id or "anonymous",
-        default=False,
-    )
-    if not allowed:
-        logger.info(f"Ignoring mode={mode} — CHAT_MODE_OPTION is disabled for user")
-        return None
-    return mode
-
-
-async def resolve_use_sdk_for_mode(
-    mode: CopilotMode | None,
-    user_id: str | None,
-    *,
-    use_claude_code_subscription: bool,
-    config_default: bool,
-) -> bool:
-    """Pick the SDK vs baseline path for a single turn.
-
-    Per-request ``mode`` wins whenever it is set (after the
-    ``CHAT_MODE_OPTION`` gate has been applied upstream).  Otherwise
-    falls back to the Claude Code subscription override, then the
-    ``COPILOT_SDK`` LaunchDarkly flag, then the config default.
-    """
-    if mode == "fast":
-        return False
-    if mode == "extended_thinking":
-        return True
-    return use_claude_code_subscription or await is_feature_enabled(
-        Flag.COPILOT_SDK,
-        user_id or "anonymous",
-        default=config_default,
-    )
-
-
 # ============ Module Entry Points ============ #
 
 # Thread-local storage for processor instances
@@ -301,17 +250,23 @@ class CoPilotProcessor:
             if config.test_mode:
                 stream_fn = stream_chat_completion_dummy
                 log.warning("Using DUMMY service (CHAT_TEST_MODE=true)")
-                effective_mode = None
             else:
-                # Enforce server-side feature-flag gate so unauthorised
-                # users cannot force a mode by crafting the request.
-                effective_mode = await resolve_effective_mode(entry.mode, entry.user_id)
-                use_sdk = await resolve_use_sdk_for_mode(
-                    effective_mode,
-                    entry.user_id,
-                    use_claude_code_subscription=config.use_claude_code_subscription,
-                    config_default=config.use_claude_agent_sdk,
-                )
+                # Per-request mode override from the frontend takes priority.
+                # 'fast' → baseline (OpenAI-compatible), 'extended_thinking' → SDK.
+                if entry.mode == "fast":
+                    use_sdk = False
+                elif entry.mode == "extended_thinking":
+                    use_sdk = True
+                else:
+                    # No mode specified — fall back to feature flag / config.
+                    use_sdk = (
+                        config.use_claude_code_subscription
+                        or await is_feature_enabled(
+                            Flag.COPILOT_SDK,
+                            entry.user_id or "anonymous",
+                            default=config.use_claude_agent_sdk,
+                        )
+                    )
                 stream_fn = (
                     sdk_service.stream_chat_completion_sdk
                     if use_sdk
@@ -319,7 +274,7 @@ class CoPilotProcessor:
                 )
                 log.info(
                     f"Using {'SDK' if use_sdk else 'baseline'} service "
-                    f"(mode={effective_mode or 'default'})"
+                    f"(mode={entry.mode or 'default'})"
                 )
 
             # Stream chat completion and publish chunks to Redis.
@@ -332,7 +287,6 @@ class CoPilotProcessor:
                 user_id=entry.user_id,
                 context=entry.context,
                 file_ids=entry.file_ids,
-                mode=effective_mode,
             )
             async for chunk in stream_registry.stream_and_publish(
                 session_id=entry.session_id,

autogpt_platform/backend/backend/copilot/executor/processor_test.py

@@ -1,175 +0,0 @@
-"""Unit tests for CoPilot mode routing logic in the processor.
-
-Tests cover the mode→service mapping:
-  - 'fast' → baseline service
-  - 'extended_thinking' → SDK service
-  - None → feature flag / config fallback
-
-as well as the ``CHAT_MODE_OPTION`` server-side gate.  The tests import
-the real production helpers from ``processor.py`` so the routing logic
-has meaningful coverage.
-"""
-
-from unittest.mock import AsyncMock, patch
-
-import pytest
-
-from backend.copilot.executor.processor import (
-    resolve_effective_mode,
-    resolve_use_sdk_for_mode,
-)
-
-
-class TestResolveUseSdkForMode:
-    """Tests for the per-request mode routing logic."""
-
-    @pytest.mark.asyncio
-    async def test_fast_mode_uses_baseline(self):
-        """mode='fast' always routes to baseline, regardless of flags."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=True),
-        ):
-            assert (
-                await resolve_use_sdk_for_mode(
-                    "fast",
-                    "user-1",
-                    use_claude_code_subscription=True,
-                    config_default=True,
-                )
-                is False
-            )
-
-    @pytest.mark.asyncio
-    async def test_extended_thinking_uses_sdk(self):
-        """mode='extended_thinking' always routes to SDK, regardless of flags."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ):
-            assert (
-                await resolve_use_sdk_for_mode(
-                    "extended_thinking",
-                    "user-1",
-                    use_claude_code_subscription=False,
-                    config_default=False,
-                )
-                is True
-            )
-
-    @pytest.mark.asyncio
-    async def test_none_mode_uses_subscription_override(self):
-        """mode=None with claude_code_subscription=True routes to SDK."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ):
-            assert (
-                await resolve_use_sdk_for_mode(
-                    None,
-                    "user-1",
-                    use_claude_code_subscription=True,
-                    config_default=False,
-                )
-                is True
-            )
-
-    @pytest.mark.asyncio
-    async def test_none_mode_uses_feature_flag(self):
-        """mode=None with feature flag enabled routes to SDK."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=True),
-        ) as flag_mock:
-            assert (
-                await resolve_use_sdk_for_mode(
-                    None,
-                    "user-1",
-                    use_claude_code_subscription=False,
-                    config_default=False,
-                )
-                is True
-            )
-            flag_mock.assert_awaited_once()
-
-    @pytest.mark.asyncio
-    async def test_none_mode_uses_config_default(self):
-        """mode=None falls back to config.use_claude_agent_sdk."""
-        # When LaunchDarkly returns the default (True), we expect SDK routing.
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=True),
-        ):
-            assert (
-                await resolve_use_sdk_for_mode(
-                    None,
-                    "user-1",
-                    use_claude_code_subscription=False,
-                    config_default=True,
-                )
-                is True
-            )
-
-    @pytest.mark.asyncio
-    async def test_none_mode_all_disabled(self):
-        """mode=None with all flags off routes to baseline."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ):
-            assert (
-                await resolve_use_sdk_for_mode(
-                    None,
-                    "user-1",
-                    use_claude_code_subscription=False,
-                    config_default=False,
-                )
-                is False
-            )
-
-
-class TestResolveEffectiveMode:
-    """Tests for the CHAT_MODE_OPTION server-side gate."""
-
-    @pytest.mark.asyncio
-    async def test_none_mode_passes_through(self):
-        """mode=None is returned as-is without a flag check."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ) as flag_mock:
-            assert await resolve_effective_mode(None, "user-1") is None
-            flag_mock.assert_not_awaited()
-
-    @pytest.mark.asyncio
-    async def test_mode_stripped_when_flag_disabled(self):
-        """When CHAT_MODE_OPTION is off, mode is dropped to None."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ):
-            assert await resolve_effective_mode("fast", "user-1") is None
-            assert await resolve_effective_mode("extended_thinking", "user-1") is None
-
-    @pytest.mark.asyncio
-    async def test_mode_preserved_when_flag_enabled(self):
-        """When CHAT_MODE_OPTION is on, the user-selected mode is preserved."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=True),
-        ):
-            assert await resolve_effective_mode("fast", "user-1") == "fast"
-            assert (
-                await resolve_effective_mode("extended_thinking", "user-1")
-                == "extended_thinking"
-            )
-
-    @pytest.mark.asyncio
-    async def test_anonymous_user_with_mode(self):
-        """Anonymous users (user_id=None) still pass through the gate."""
-        with patch(
-            "backend.copilot.executor.processor.is_feature_enabled",
-            new=AsyncMock(return_value=False),
-        ) as flag_mock:
-            assert await resolve_effective_mode("fast", None) is None
-            flag_mock.assert_awaited_once()

autogpt_platform/backend/backend/copilot/executor/utils.py

@@ -6,10 +6,10 @@ Defines two exchanges and queues following the graph executor pattern:
 """
 
 import logging
+from typing import Literal
 
 from pydantic import BaseModel
 
-from backend.copilot.config import CopilotMode
 from backend.data.rabbitmq import Exchange, ExchangeType, Queue, RabbitMQConfig
 from backend.util.logging import TruncatedLogger, is_structured_logging_enabled
 
@@ -157,7 +157,7 @@ class CoPilotExecutionEntry(BaseModel):
     file_ids: list[str] | None = None
     """Workspace file IDs attached to the user's message"""
 
-    mode: CopilotMode | None = None
+    mode: Literal["fast", "extended_thinking"] | None = None
     """Autopilot mode override: 'fast' or 'extended_thinking'. None = server default."""
 
 
@@ -179,7 +179,7 @@ async def enqueue_copilot_turn(
     is_user_message: bool = True,
     context: dict[str, str] | None = None,
     file_ids: list[str] | None = None,
-    mode: CopilotMode | None = None,
+    mode: Literal["fast", "extended_thinking"] | None = None,
 ) -> None:
     """Enqueue a CoPilot task for processing by the executor service.
 

autogpt_platform/backend/backend/copilot/executor/utils_test.py

@@ -1,123 +0,0 @@
-"""Tests for CoPilot executor utils (queue config, message models, logging)."""
-
-from backend.copilot.executor.utils import (
-    COPILOT_EXECUTION_EXCHANGE,
-    COPILOT_EXECUTION_QUEUE_NAME,
-    COPILOT_EXECUTION_ROUTING_KEY,
-    CancelCoPilotEvent,
-    CoPilotExecutionEntry,
-    CoPilotLogMetadata,
-    create_copilot_queue_config,
-)
-
-
-class TestCoPilotExecutionEntry:
-    def test_basic_fields(self):
-        entry = CoPilotExecutionEntry(
-            session_id="s1",
-            user_id="u1",
-            message="hello",
-        )
-        assert entry.session_id == "s1"
-        assert entry.user_id == "u1"
-        assert entry.message == "hello"
-        assert entry.is_user_message is True
-        assert entry.mode is None
-        assert entry.context is None
-        assert entry.file_ids is None
-
-    def test_mode_field(self):
-        entry = CoPilotExecutionEntry(
-            session_id="s1",
-            user_id="u1",
-            message="test",
-            mode="fast",
-        )
-        assert entry.mode == "fast"
-
-        entry2 = CoPilotExecutionEntry(
-            session_id="s1",
-            user_id="u1",
-            message="test",
-            mode="extended_thinking",
-        )
-        assert entry2.mode == "extended_thinking"
-
-    def test_optional_fields(self):
-        entry = CoPilotExecutionEntry(
-            session_id="s1",
-            user_id="u1",
-            message="test",
-            turn_id="t1",
-            context={"url": "https://example.com"},
-            file_ids=["f1", "f2"],
-            is_user_message=False,
-        )
-        assert entry.turn_id == "t1"
-        assert entry.context == {"url": "https://example.com"}
-        assert entry.file_ids == ["f1", "f2"]
-        assert entry.is_user_message is False
-
-    def test_serialization_roundtrip(self):
-        entry = CoPilotExecutionEntry(
-            session_id="s1",
-            user_id="u1",
-            message="hello",
-            mode="fast",
-        )
-        json_str = entry.model_dump_json()
-        restored = CoPilotExecutionEntry.model_validate_json(json_str)
-        assert restored == entry
-
-
-class TestCancelCoPilotEvent:
-    def test_basic(self):
-        event = CancelCoPilotEvent(session_id="s1")
-        assert event.session_id == "s1"
-
-    def test_serialization(self):
-        event = CancelCoPilotEvent(session_id="s1")
-        restored = CancelCoPilotEvent.model_validate_json(event.model_dump_json())
-        assert restored.session_id == "s1"
-
-
-class TestCreateCopilotQueueConfig:
-    def test_returns_valid_config(self):
-        config = create_copilot_queue_config()
-        assert len(config.exchanges) == 2
-        assert len(config.queues) == 2
-
-    def test_execution_queue_properties(self):
-        config = create_copilot_queue_config()
-        exec_queue = next(
-            q for q in config.queues if q.name == COPILOT_EXECUTION_QUEUE_NAME
-        )
-        assert exec_queue.durable is True
-        assert exec_queue.exchange == COPILOT_EXECUTION_EXCHANGE
-        assert exec_queue.routing_key == COPILOT_EXECUTION_ROUTING_KEY
-
-    def test_cancel_queue_uses_fanout(self):
-        config = create_copilot_queue_config()
-        cancel_queue = next(
-            q for q in config.queues if q.name != COPILOT_EXECUTION_QUEUE_NAME
-        )
-        assert cancel_queue.exchange is not None
-        assert cancel_queue.exchange.type.value == "fanout"
-
-
-class TestCoPilotLogMetadata:
-    def test_creates_logger_with_metadata(self):
-        import logging
-
-        base_logger = logging.getLogger("test")
-        log = CoPilotLogMetadata(base_logger, session_id="s1", user_id="u1")
-        assert log is not None
-
-    def test_filters_none_values(self):
-        import logging
-
-        base_logger = logging.getLogger("test")
-        log = CoPilotLogMetadata(
-            base_logger, session_id="s1", user_id=None, turn_id="t1"
-        )
-        assert log is not None

autogpt_platform/backend/backend/copilot/model.py

@@ -64,7 +64,6 @@ class ChatMessage(BaseModel):
     refusal: str | None = None
     tool_calls: list[dict] | None = None
     function_call: dict | None = None
-    sequence: int | None = None
     duration_ms: int | None = None
 
     @staticmethod
@@ -78,54 +77,10 @@ class ChatMessage(BaseModel):
             refusal=prisma_message.refusal,
             tool_calls=_parse_json_field(prisma_message.toolCalls),
             function_call=_parse_json_field(prisma_message.functionCall),
-            sequence=prisma_message.sequence,
             duration_ms=prisma_message.durationMs,
         )
 
 
-def is_message_duplicate(
-    messages: list[ChatMessage],
-    role: str,
-    content: str,
-) -> bool:
-    """Check whether *content* is already present in the current pending turn.
-
-    Only inspects trailing messages that share the given *role* (i.e. the
-    current turn). This ensures legitimately repeated messages across different
-    turns are not suppressed, while same-turn duplicates from stale cache are
-    still caught.
-    """
-    for m in reversed(messages):
-        if m.role == role:
-            if m.content == content:
-                return True
-        else:
-            break
-    return False
-
-
-def maybe_append_user_message(
-    session: "ChatSession",
-    message: str | None,
-    is_user_message: bool,
-) -> bool:
-    """Append a user/assistant message to the session if not already present.
-
-    The route handler already persists the user message before enqueueing,
-    so we check trailing same-role messages to avoid re-appending when the
-    session cache is slightly stale.
-
-    Returns True if the message was appended, False if skipped.
-    """
-    if not message:
-        return False
-    role = "user" if is_user_message else "assistant"
-    if is_message_duplicate(session.messages, role, message):
-        return False
-    session.messages.append(ChatMessage(role=role, content=message))
-    return True
-
-
 class Usage(BaseModel):
     prompt_tokens: int
     completion_tokens: int

autogpt_platform/backend/backend/copilot/model_test.py

@@ -17,8 +17,6 @@ from .model import (
     ChatSession,
     Usage,
     get_chat_session,
-    is_message_duplicate,
-    maybe_append_user_message,
     upsert_chat_session,
 )
 
@@ -426,151 +424,3 @@ async def test_concurrent_saves_collision_detection(setup_test_user, test_user_i
     assert "Streaming message 1" in contents
     assert "Streaming message 2" in contents
     assert "Callback result" in contents
-
-
-# --------------------------------------------------------------------------- #
-#  is_message_duplicate                                                        #
-# --------------------------------------------------------------------------- #
-
-
-def test_duplicate_detected_in_trailing_same_role():
-    """Duplicate user message at the tail is detected."""
-    msgs = [
-        ChatMessage(role="user", content="hello"),
-        ChatMessage(role="assistant", content="hi there"),
-        ChatMessage(role="user", content="yes"),
-    ]
-    assert is_message_duplicate(msgs, "user", "yes") is True
-
-
-def test_duplicate_not_detected_across_turns():
-    """Same text in a previous turn (separated by assistant) is NOT a duplicate."""
-    msgs = [
-        ChatMessage(role="user", content="yes"),
-        ChatMessage(role="assistant", content="ok"),
-    ]
-    assert is_message_duplicate(msgs, "user", "yes") is False
-
-
-def test_no_duplicate_on_empty_messages():
-    """Empty message list never reports a duplicate."""
-    assert is_message_duplicate([], "user", "hello") is False
-
-
-def test_no_duplicate_when_content_differs():
-    """Different content in the trailing same-role block is not a duplicate."""
-    msgs = [
-        ChatMessage(role="assistant", content="response"),
-        ChatMessage(role="user", content="first message"),
-    ]
-    assert is_message_duplicate(msgs, "user", "second message") is False
-
-
-def test_duplicate_with_multiple_trailing_same_role():
-    """Detects duplicate among multiple consecutive same-role messages."""
-    msgs = [
-        ChatMessage(role="assistant", content="response"),
-        ChatMessage(role="user", content="msg1"),
-        ChatMessage(role="user", content="msg2"),
-    ]
-    assert is_message_duplicate(msgs, "user", "msg1") is True
-    assert is_message_duplicate(msgs, "user", "msg2") is True
-    assert is_message_duplicate(msgs, "user", "msg3") is False
-
-
-def test_duplicate_check_for_assistant_role():
-    """Works correctly when checking assistant role too."""
-    msgs = [
-        ChatMessage(role="user", content="hi"),
-        ChatMessage(role="assistant", content="hello"),
-        ChatMessage(role="assistant", content="how can I help?"),
-    ]
-    assert is_message_duplicate(msgs, "assistant", "hello") is True
-    assert is_message_duplicate(msgs, "assistant", "new response") is False
-
-
-def test_no_false_positive_when_content_is_none():
-    """Messages with content=None in the trailing block do not match."""
-    msgs = [
-        ChatMessage(role="user", content=None),
-        ChatMessage(role="user", content="hello"),
-    ]
-    assert is_message_duplicate(msgs, "user", "hello") is True
-    # None-content message should not match any string
-    msgs2 = [
-        ChatMessage(role="user", content=None),
-    ]
-    assert is_message_duplicate(msgs2, "user", "hello") is False
-
-
-def test_all_same_role_messages():
-    """When all messages share the same role, the entire list is scanned."""
-    msgs = [
-        ChatMessage(role="user", content="first"),
-        ChatMessage(role="user", content="second"),
-        ChatMessage(role="user", content="third"),
-    ]
-    assert is_message_duplicate(msgs, "user", "first") is True
-    assert is_message_duplicate(msgs, "user", "new") is False
-
-
-# --------------------------------------------------------------------------- #
-#  maybe_append_user_message                                                   #
-# --------------------------------------------------------------------------- #
-
-
-def test_maybe_append_user_message_appends_new():
-    """A new user message is appended and returns True."""
-    session = ChatSession.new(user_id="u", dry_run=False)
-    session.messages = [
-        ChatMessage(role="assistant", content="hello"),
-    ]
-    result = maybe_append_user_message(session, "new msg", is_user_message=True)
-    assert result is True
-    assert len(session.messages) == 2
-    assert session.messages[-1].role == "user"
-    assert session.messages[-1].content == "new msg"
-
-
-def test_maybe_append_user_message_skips_duplicate():
-    """A duplicate user message is skipped and returns False."""
-    session = ChatSession.new(user_id="u", dry_run=False)
-    session.messages = [
-        ChatMessage(role="assistant", content="hello"),
-        ChatMessage(role="user", content="dup"),
-    ]
-    result = maybe_append_user_message(session, "dup", is_user_message=True)
-    assert result is False
-    assert len(session.messages) == 2
-
-
-def test_maybe_append_user_message_none_message():
-    """None/empty message returns False without appending."""
-    session = ChatSession.new(user_id="u", dry_run=False)
-    assert maybe_append_user_message(session, None, is_user_message=True) is False
-    assert maybe_append_user_message(session, "", is_user_message=True) is False
-    assert len(session.messages) == 0
-
-
-def test_maybe_append_assistant_message():
-    """Works for assistant role when is_user_message=False."""
-    session = ChatSession.new(user_id="u", dry_run=False)
-    session.messages = [
-        ChatMessage(role="user", content="hi"),
-    ]
-    result = maybe_append_user_message(session, "response", is_user_message=False)
-    assert result is True
-    assert session.messages[-1].role == "assistant"
-    assert session.messages[-1].content == "response"
-
-
-def test_maybe_append_assistant_skips_duplicate():
-    """Duplicate assistant message is skipped."""
-    session = ChatSession.new(user_id="u", dry_run=False)
-    session.messages = [
-        ChatMessage(role="user", content="hi"),
-        ChatMessage(role="assistant", content="dup"),
-    ]
-    result = maybe_append_user_message(session, "dup", is_user_message=False)
-    assert result is False
-    assert len(session.messages) == 2

autogpt_platform/backend/backend/copilot/rate_limit_test.py

@@ -13,21 +13,12 @@ from .rate_limit import (
     RateLimitExceeded,
     SubscriptionTier,
     UsageWindow,
-    _daily_key,
-    _daily_reset_time,
-    _weekly_key,
-    _weekly_reset_time,
-    acquire_reset_lock,
     check_rate_limit,
-    get_daily_reset_count,
     get_global_rate_limits,
     get_usage_status,
     get_user_tier,
-    increment_daily_reset_count,
     record_token_usage,
-    release_reset_lock,
     reset_daily_usage,
-    reset_user_usage,
     set_user_tier,
 )
 
@@ -1219,205 +1210,3 @@ class TestTierLimitsEnforced:
             assert daily == biz_daily  # 20x
             # Should NOT raise — usage is within the BUSINESS tier allowance
             await check_rate_limit(_USER, daily, weekly)
-
-
-# ---------------------------------------------------------------------------
-# Private key/reset helpers
-# ---------------------------------------------------------------------------
-
-
-class TestKeyHelpers:
-    def test_daily_key_format(self):
-        now = datetime(2026, 4, 3, 12, 0, 0, tzinfo=UTC)
-        key = _daily_key("user-1", now=now)
-        assert "daily" in key
-        assert "user-1" in key
-        assert "2026-04-03" in key
-
-    def test_daily_key_defaults_to_now(self):
-        key = _daily_key("user-1")
-        assert "daily" in key
-        assert "user-1" in key
-
-    def test_weekly_key_format(self):
-        now = datetime(2026, 4, 3, 12, 0, 0, tzinfo=UTC)
-        key = _weekly_key("user-1", now=now)
-        assert "weekly" in key
-        assert "user-1" in key
-        assert "2026-W" in key
-
-    def test_weekly_key_defaults_to_now(self):
-        key = _weekly_key("user-1")
-        assert "weekly" in key
-
-    def test_daily_reset_time_is_next_midnight(self):
-        now = datetime(2026, 4, 3, 15, 30, 0, tzinfo=UTC)
-        reset = _daily_reset_time(now=now)
-        assert reset == datetime(2026, 4, 4, 0, 0, 0, tzinfo=UTC)
-
-    def test_daily_reset_time_defaults_to_now(self):
-        reset = _daily_reset_time()
-        assert reset.hour == 0
-        assert reset.minute == 0
-
-    def test_weekly_reset_time_is_next_monday(self):
-        # 2026-04-03 is a Friday
-        now = datetime(2026, 4, 3, 15, 30, 0, tzinfo=UTC)
-        reset = _weekly_reset_time(now=now)
-        assert reset.weekday() == 0  # Monday
-        assert reset == datetime(2026, 4, 6, 0, 0, 0, tzinfo=UTC)
-
-    def test_weekly_reset_time_defaults_to_now(self):
-        reset = _weekly_reset_time()
-        assert reset.weekday() == 0  # Monday
-
-
-# ---------------------------------------------------------------------------
-# acquire_reset_lock / release_reset_lock
-# ---------------------------------------------------------------------------
-
-
-class TestResetLock:
-    @pytest.mark.asyncio
-    async def test_acquire_lock_success(self):
-        mock_redis = AsyncMock()
-        mock_redis.set = AsyncMock(return_value=True)
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            result = await acquire_reset_lock("user-1")
-        assert result is True
-
-    @pytest.mark.asyncio
-    async def test_acquire_lock_already_held(self):
-        mock_redis = AsyncMock()
-        mock_redis.set = AsyncMock(return_value=False)
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            result = await acquire_reset_lock("user-1")
-        assert result is False
-
-    @pytest.mark.asyncio
-    async def test_acquire_lock_redis_unavailable(self):
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async",
-            side_effect=RedisError("down"),
-        ):
-            result = await acquire_reset_lock("user-1")
-        assert result is False
-
-    @pytest.mark.asyncio
-    async def test_release_lock_success(self):
-        mock_redis = AsyncMock()
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            await release_reset_lock("user-1")
-        mock_redis.delete.assert_called_once()
-
-    @pytest.mark.asyncio
-    async def test_release_lock_redis_unavailable(self):
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async",
-            side_effect=RedisError("down"),
-        ):
-            # Should not raise
-            await release_reset_lock("user-1")
-
-
-# ---------------------------------------------------------------------------
-# get_daily_reset_count / increment_daily_reset_count
-# ---------------------------------------------------------------------------
-
-
-class TestDailyResetCount:
-    @pytest.mark.asyncio
-    async def test_get_count_returns_value(self):
-        mock_redis = AsyncMock()
-        mock_redis.get = AsyncMock(return_value="3")
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            count = await get_daily_reset_count("user-1")
-        assert count == 3
-
-    @pytest.mark.asyncio
-    async def test_get_count_returns_zero_when_no_key(self):
-        mock_redis = AsyncMock()
-        mock_redis.get = AsyncMock(return_value=None)
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            count = await get_daily_reset_count("user-1")
-        assert count == 0
-
-    @pytest.mark.asyncio
-    async def test_get_count_returns_none_when_redis_unavailable(self):
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async",
-            side_effect=RedisError("down"),
-        ):
-            count = await get_daily_reset_count("user-1")
-        assert count is None
-
-    @pytest.mark.asyncio
-    async def test_increment_count(self):
-        mock_pipe = MagicMock()
-        mock_pipe.incr = MagicMock()
-        mock_pipe.expire = MagicMock()
-        mock_pipe.execute = AsyncMock()
-
-        mock_redis = AsyncMock()
-        mock_redis.pipeline = MagicMock(return_value=mock_pipe)
-
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            await increment_daily_reset_count("user-1")
-        mock_pipe.incr.assert_called_once()
-        mock_pipe.expire.assert_called_once()
-
-    @pytest.mark.asyncio
-    async def test_increment_count_redis_unavailable(self):
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async",
-            side_effect=RedisError("down"),
-        ):
-            # Should not raise
-            await increment_daily_reset_count("user-1")
-
-
-# ---------------------------------------------------------------------------
-# reset_user_usage
-# ---------------------------------------------------------------------------
-
-
-class TestResetUserUsage:
-    @pytest.mark.asyncio
-    async def test_resets_daily_key(self):
-        mock_redis = AsyncMock()
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            await reset_user_usage("user-1")
-        mock_redis.delete.assert_called_once()
-
-    @pytest.mark.asyncio
-    async def test_resets_daily_and_weekly(self):
-        mock_redis = AsyncMock()
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async", return_value=mock_redis
-        ):
-            await reset_user_usage("user-1", reset_weekly=True)
-        args = mock_redis.delete.call_args[0]
-        assert len(args) == 2  # both daily and weekly keys
-
-    @pytest.mark.asyncio
-    async def test_raises_on_redis_failure(self):
-        with patch(
-            "backend.copilot.rate_limit.get_redis_async",
-            side_effect=RedisError("down"),
-        ):
-            with pytest.raises(RedisError):
-                await reset_user_usage("user-1")

autogpt_platform/backend/backend/copilot/sdk/p0_guardrails_test.py

@@ -0,0 +1,391 @@
+"""Tests for P0 guardrails: _resolve_fallback_model, security env vars, TMPDIR."""
+
+from unittest.mock import patch
+
+import pytest
+from pydantic import ValidationError
+
+from backend.copilot.config import ChatConfig
+from backend.copilot.constants import is_transient_api_error
+
+
+def _make_config(**overrides) -> ChatConfig:
+    """Create a ChatConfig with safe defaults, applying *overrides*."""
+    defaults = {
+        "use_claude_code_subscription": False,
+        "use_openrouter": False,
+        "api_key": None,
+        "base_url": None,
+    }
+    defaults.update(overrides)
+    return ChatConfig(**defaults)
+
+
+# ---------------------------------------------------------------------------
+# _resolve_fallback_model
+# ---------------------------------------------------------------------------
+
+_SVC = "backend.copilot.sdk.service"
+
+
+class TestResolveFallbackModel:
+    """Provider-aware fallback model resolution."""
+
+    def test_returns_none_when_empty(self):
+        cfg = _make_config(claude_agent_fallback_model="")
+        with patch(f"{_SVC}.config", cfg):
+            from backend.copilot.sdk.service import _resolve_fallback_model
+
+            assert _resolve_fallback_model() is None
+
+    def test_strips_provider_prefix(self):
+        """OpenRouter-style 'anthropic/claude-sonnet-4-...' is stripped."""
+        cfg = _make_config(
+            claude_agent_fallback_model="anthropic/claude-sonnet-4-20250514",
+            use_openrouter=True,
+            api_key="sk-test",
+            base_url="https://openrouter.ai/api/v1",
+        )
+        with patch(f"{_SVC}.config", cfg):
+            from backend.copilot.sdk.service import _resolve_fallback_model
+
+            result = _resolve_fallback_model()
+
+        assert result == "claude-sonnet-4-20250514"
+        assert "/" not in result
+
+    def test_dots_replaced_for_direct_anthropic(self):
+        """Direct Anthropic requires hyphen-separated versions."""
+        cfg = _make_config(
+            claude_agent_fallback_model="claude-sonnet-4.5-20250514",
+            use_openrouter=False,
+        )
+        with patch(f"{_SVC}.config", cfg):
+            from backend.copilot.sdk.service import _resolve_fallback_model
+
+            result = _resolve_fallback_model()
+
+        assert result is not None
+        assert "." not in result
+        assert result == "claude-sonnet-4-5-20250514"
+
+    def test_dots_preserved_for_openrouter(self):
+        """OpenRouter uses dot-separated versions — don't normalise."""
+        cfg = _make_config(
+            claude_agent_fallback_model="claude-sonnet-4.5-20250514",
+            use_openrouter=True,
+            api_key="sk-test",
+            base_url="https://openrouter.ai/api/v1",
+        )
+        with patch(f"{_SVC}.config", cfg):
+            from backend.copilot.sdk.service import _resolve_fallback_model
+
+            result = _resolve_fallback_model()
+
+        assert result == "claude-sonnet-4.5-20250514"
+
+    def test_default_value(self):
+        """Default fallback model resolves to a valid string."""
+        cfg = _make_config()
+        with patch(f"{_SVC}.config", cfg):
+            from backend.copilot.sdk.service import _resolve_fallback_model
+
+            result = _resolve_fallback_model()
+
+        assert result is not None
+        assert "sonnet" in result.lower() or "claude" in result.lower()
+
+
+# ---------------------------------------------------------------------------
+# Security & isolation env vars
+# ---------------------------------------------------------------------------
+
+
+class TestSecurityEnvVars:
+    """Verify the env-var contract in the service module.
+
+    The production code sets CLAUDE_CODE_TMPDIR and security env vars
+    inline after ``build_sdk_env()`` returns.  We grep for these string
+    literals in ``service.py`` to ensure they aren't accidentally removed.
+    """
+
+    _SERVICE_PATH = "autogpt_platform/backend/backend/copilot/sdk/service.py"
+
+    @staticmethod
+    def _read_service_source() -> str:
+        import pathlib
+
+        # Walk up from this test file to the repo root
+        repo = pathlib.Path(__file__).resolve().parents[5]
+        return (repo / TestSecurityEnvVars._SERVICE_PATH).read_text()
+
+    def test_tmpdir_env_var_present_in_source(self):
+        """CLAUDE_CODE_TMPDIR must be set when sdk_cwd is provided."""
+        src = self._read_service_source()
+        assert 'sdk_env["CLAUDE_CODE_TMPDIR"]' in src
+
+    def test_home_not_overridden_in_source(self):
+        """HOME must NOT be overridden — would break git/ssh/npm."""
+        src = self._read_service_source()
+        assert 'sdk_env["HOME"]' not in src
+
+    def test_security_env_vars_present_in_source(self):
+        """All four security env vars must be set in the service module."""
+        src = self._read_service_source()
+        for var in (
+            "CLAUDE_CODE_DISABLE_CLAUDE_MDS",
+            "CLAUDE_CODE_SKIP_PROMPT_HISTORY",
+            "CLAUDE_CODE_DISABLE_AUTO_MEMORY",
+            "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC",
+        ):
+            assert var in src, f"{var} not found in service.py"
+
+
+# ---------------------------------------------------------------------------
+# Config defaults
+# ---------------------------------------------------------------------------
+
+
+class TestConfigDefaults:
+    """Verify ChatConfig P0 fields have correct defaults."""
+
+    def test_fallback_model_default(self):
+        cfg = _make_config()
+        assert cfg.claude_agent_fallback_model
+        assert "sonnet" in cfg.claude_agent_fallback_model.lower()
+
+    def test_max_turns_default(self):
+        cfg = _make_config()
+        assert cfg.claude_agent_max_turns == 50
+
+    def test_max_budget_usd_default(self):
+        cfg = _make_config()
+        assert cfg.claude_agent_max_budget_usd == 5.0
+
+    def test_max_transient_retries_default(self):
+        cfg = _make_config()
+        assert cfg.claude_agent_max_transient_retries == 3
+
+
+# ---------------------------------------------------------------------------
+# build_sdk_env — all 3 auth modes
+# ---------------------------------------------------------------------------
+
+_ENV = "backend.copilot.sdk.env"
+
+
+class TestBuildSdkEnv:
+    """Verify build_sdk_env returns correct dicts for each auth mode."""
+
+    def test_subscription_mode_clears_keys(self):
+        """Mode 1: subscription clears API key / auth token / base URL."""
+        cfg = _make_config(use_claude_code_subscription=True)
+        with (
+            patch(f"{_ENV}.config", cfg),
+            patch(f"{_ENV}.validate_subscription"),
+        ):
+            from backend.copilot.sdk.env import build_sdk_env
+
+            env = build_sdk_env(session_id="s1", user_id="u1")
+
+        assert env["ANTHROPIC_API_KEY"] == ""
+        assert env["ANTHROPIC_AUTH_TOKEN"] == ""
+        assert env["ANTHROPIC_BASE_URL"] == ""
+
+    def test_direct_anthropic_returns_empty_dict(self):
+        """Mode 2: direct Anthropic returns {} (inherits from parent env)."""
+        cfg = _make_config(
+            use_claude_code_subscription=False,
+            use_openrouter=False,
+        )
+        with patch(f"{_ENV}.config", cfg):
+            from backend.copilot.sdk.env import build_sdk_env
+
+            env = build_sdk_env()
+
+        assert env == {}
+
+    def test_openrouter_sets_base_url_and_auth(self):
+        """Mode 3: OpenRouter sets base URL, auth token, and clears API key."""
+        cfg = _make_config(
+            use_claude_code_subscription=False,
+            use_openrouter=True,
+            api_key="sk-or-test",
+            base_url="https://openrouter.ai/api/v1",
+        )
+        with patch(f"{_ENV}.config", cfg):
+            from backend.copilot.sdk.env import build_sdk_env
+
+            env = build_sdk_env(session_id="sess-1", user_id="user-1")
+
+        assert env["ANTHROPIC_BASE_URL"] == "https://openrouter.ai/api"
+        assert env["ANTHROPIC_AUTH_TOKEN"] == "sk-or-test"
+        assert env["ANTHROPIC_API_KEY"] == ""
+        assert "x-session-id: sess-1" in env["ANTHROPIC_CUSTOM_HEADERS"]
+        assert "x-user-id: user-1" in env["ANTHROPIC_CUSTOM_HEADERS"]
+
+    def test_openrouter_no_headers_when_ids_empty(self):
+        """Mode 3: No custom headers when session_id/user_id are not given."""
+        cfg = _make_config(
+            use_claude_code_subscription=False,
+            use_openrouter=True,
+            api_key="sk-or-test",
+            base_url="https://openrouter.ai/api/v1",
+        )
+        with patch(f"{_ENV}.config", cfg):
+            from backend.copilot.sdk.env import build_sdk_env
+
+            env = build_sdk_env()
+
+        assert "ANTHROPIC_CUSTOM_HEADERS" not in env
+
+    def test_all_modes_return_mutable_dict(self):
+        """build_sdk_env must return a mutable dict (not None) so callers
+        can add security env vars like CLAUDE_CODE_TMPDIR."""
+        for cfg in (
+            _make_config(use_claude_code_subscription=True),
+            _make_config(use_openrouter=False),
+            _make_config(
+                use_openrouter=True,
+                api_key="k",
+                base_url="https://openrouter.ai/api/v1",
+            ),
+        ):
+            with (
+                patch(f"{_ENV}.config", cfg),
+                patch(f"{_ENV}.validate_subscription"),
+            ):
+                from backend.copilot.sdk.env import build_sdk_env
+
+                env = build_sdk_env()
+
+            assert isinstance(env, dict)
+            env["CLAUDE_CODE_TMPDIR"] = "/tmp/test"
+            assert env["CLAUDE_CODE_TMPDIR"] == "/tmp/test"
+
+
+# ---------------------------------------------------------------------------
+# is_transient_api_error
+# ---------------------------------------------------------------------------
+
+
+class TestIsTransientApiError:
+    """Verify that is_transient_api_error detects all transient patterns."""
+
+    @pytest.mark.parametrize(
+        "error_text",
+        [
+            "socket connection was closed unexpectedly",
+            "ECONNRESET",
+            "connection was forcibly closed",
+            "network socket disconnected",
+        ],
+    )
+    def test_connection_level_errors(self, error_text: str):
+        assert is_transient_api_error(error_text)
+
+    @pytest.mark.parametrize(
+        "error_text",
+        [
+            "rate limit exceeded",
+            "rate_limit_error",
+            "Too Many Requests",
+            "status code 429",
+        ],
+    )
+    def test_429_rate_limit_errors(self, error_text: str):
+        assert is_transient_api_error(error_text)
+
+    @pytest.mark.parametrize(
+        "error_text",
+        [
+            "API is overloaded",
+            "Internal Server Error",
+            "Bad Gateway",
+            "Service Unavailable",
+            "Gateway Timeout",
+            "status code 529",
+            "status code 500",
+            "status code 502",
+            "status code 503",
+            "status code 504",
+        ],
+    )
+    def test_5xx_server_errors(self, error_text: str):
+        assert is_transient_api_error(error_text)
+
+    @pytest.mark.parametrize(
+        "error_text",
+        [
+            "invalid_api_key",
+            "Authentication failed",
+            "prompt is too long",
+            "model not found",
+            "",
+        ],
+    )
+    def test_non_transient_errors(self, error_text: str):
+        assert not is_transient_api_error(error_text)
+
+    def test_case_insensitive(self):
+        assert is_transient_api_error("SOCKET CONNECTION WAS CLOSED UNEXPECTEDLY")
+        assert is_transient_api_error("econnreset")
+
+
+# ---------------------------------------------------------------------------
+# Config validators for max_turns / max_budget_usd
+# ---------------------------------------------------------------------------
+
+
+class TestConfigValidators:
+    """Verify ge/le bounds on max_turns and max_budget_usd."""
+
+    def test_max_turns_rejects_zero(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_turns=0)
+
+    def test_max_turns_rejects_negative(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_turns=-1)
+
+    def test_max_turns_rejects_above_500(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_turns=501)
+
+    def test_max_turns_accepts_boundary_values(self):
+        cfg_low = _make_config(claude_agent_max_turns=1)
+        assert cfg_low.claude_agent_max_turns == 1
+        cfg_high = _make_config(claude_agent_max_turns=500)
+        assert cfg_high.claude_agent_max_turns == 500
+
+    def test_max_budget_rejects_zero(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_budget_usd=0.0)
+
+    def test_max_budget_rejects_negative(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_budget_usd=-1.0)
+
+    def test_max_budget_rejects_above_100(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_budget_usd=100.01)
+
+    def test_max_budget_accepts_boundary_values(self):
+        cfg_low = _make_config(claude_agent_max_budget_usd=0.01)
+        assert cfg_low.claude_agent_max_budget_usd == 0.01
+        cfg_high = _make_config(claude_agent_max_budget_usd=100.0)
+        assert cfg_high.claude_agent_max_budget_usd == 100.0
+
+    def test_max_transient_retries_rejects_negative(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_transient_retries=-1)
+
+    def test_max_transient_retries_rejects_above_10(self):
+        with pytest.raises(ValidationError):
+            _make_config(claude_agent_max_transient_retries=11)
+
+    def test_max_transient_retries_accepts_boundary_values(self):
+        cfg_low = _make_config(claude_agent_max_transient_retries=0)
+        assert cfg_low.claude_agent_max_transient_retries == 0
+        cfg_high = _make_config(claude_agent_max_transient_retries=10)
+        assert cfg_high.claude_agent_max_transient_retries == 10

autogpt_platform/backend/backend/copilot/sdk/prompt_too_long_test.py

@@ -8,19 +8,20 @@ from uuid import uuid4
 
 import pytest
 
-from backend.copilot.transcript import (
-    _flatten_assistant_content,
-    _flatten_tool_result_content,
-    _messages_to_transcript,
-    _run_compression,
-    _transcript_to_messages,
-)
 from backend.util import json
 from backend.util.prompt import CompressResult
 
 from .conftest import build_test_transcript as _build_transcript
 from .service import _friendly_error_text, _is_prompt_too_long
-from .transcript import compact_transcript, validate_transcript
+from .transcript import (
+    _flatten_assistant_content,
+    _flatten_tool_result_content,
+    _messages_to_transcript,
+    _run_compression,
+    _transcript_to_messages,
+    compact_transcript,
+    validate_transcript,
+)
 
 # ---------------------------------------------------------------------------
 # _flatten_assistant_content

autogpt_platform/backend/backend/copilot/sdk/response_adapter_test.py

@@ -260,13 +260,13 @@ def test_result_error_emits_error_and_finish():
         is_error=True,
         num_turns=0,
         session_id="s1",
-        result="API rate limited",
+        result="Invalid API key provided",
     )
     results = adapter.convert_message(msg)
     # No step was open, so no FinishStep — just Error + Finish
     assert len(results) == 2
     assert isinstance(results[0], StreamError)
-    assert "API rate limited" in results[0].errorText
+    assert "Invalid API key provided" in results[0].errorText
     assert isinstance(results[1], StreamFinish)
 
 

autogpt_platform/backend/backend/copilot/sdk/retry_scenarios_test.py

@@ -26,17 +26,18 @@ from unittest.mock import AsyncMock, MagicMock, patch
 
 import pytest
 
-from backend.copilot.transcript import (
-    _flatten_assistant_content,
-    _flatten_tool_result_content,
-    _messages_to_transcript,
-    _transcript_to_messages,
-)
 from backend.util import json
 
 from .conftest import build_test_transcript as _build_transcript
 from .service import _MAX_STREAM_ATTEMPTS, _reduce_context
-from .transcript import compact_transcript, validate_transcript
+from .transcript import (
+    _flatten_assistant_content,
+    _flatten_tool_result_content,
+    _messages_to_transcript,
+    _transcript_to_messages,
+    compact_transcript,
+    validate_transcript,
+)
 from .transcript_builder import TranscriptBuilder
 
 # ---------------------------------------------------------------------------
@@ -1404,9 +1405,9 @@ class TestStreamChatCompletionRetryIntegration:
                 events.append(event)
 
         # Should NOT retry — only 1 attempt for auth errors
-        assert (
-            attempt_count[0] == 1
-        ), f"Expected 1 attempt (no retry for auth error), got {attempt_count[0]}"
+        assert attempt_count[0] == 1, (
+            f"Expected 1 attempt (no retry for auth error), " f"got {attempt_count[0]}"
+        )
         errors = [e for e in events if isinstance(e, StreamError)]
         assert errors, "Expected StreamError"
         assert errors[0].code == "sdk_stream_error"

autogpt_platform/backend/backend/copilot/sdk/sdk_compat_test.py

@@ -105,6 +105,10 @@ def test_agent_options_accepts_all_our_fields():
         "env",
         "resume",
         "max_buffer_size",
+        "stderr",
+        "fallback_model",
+        "max_turns",
+        "max_budget_usd",
     ]
     sig = inspect.signature(ClaudeAgentOptions)
     for field in fields_we_use:

autogpt_platform/backend/backend/copilot/sdk/service.py

@@ -33,24 +33,12 @@ from pydantic import BaseModel
 
 from backend.copilot.context import get_workspace_manager
 from backend.copilot.permissions import apply_tool_permissions
-from backend.copilot.rate_limit import get_user_tier
-from backend.copilot.transcript import (
-    _run_compression,
-    cleanup_stale_project_dirs,
-    compact_transcript,
-    download_transcript,
-    read_compacted_entries,
-    upload_transcript,
-    validate_transcript,
-    write_transcript_to_tempfile,
-)
-from backend.copilot.transcript_builder import TranscriptBuilder
 from backend.data.redis_client import get_redis_async
 from backend.executor.cluster_lock import AsyncClusterLock
 from backend.util.exceptions import NotFoundError
 from backend.util.settings import Settings
 
-from ..config import ChatConfig, CopilotMode
+from ..config import ChatConfig
 from ..constants import (
     COPILOT_ERROR_PREFIX,
     COPILOT_RETRYABLE_ERROR_PREFIX,
@@ -63,7 +51,6 @@ from ..model import (
     ChatMessage,
     ChatSession,
     get_chat_session,
-    maybe_append_user_message,
     update_session_title,
     upsert_chat_session,
 )
@@ -105,6 +92,17 @@ from .tool_adapter import (
     set_execution_context,
     wait_for_stash,
 )
+from .transcript import (
+    _run_compression,
+    cleanup_stale_project_dirs,
+    compact_transcript,
+    download_transcript,
+    read_compacted_entries,
+    upload_transcript,
+    validate_transcript,
+    write_transcript_to_tempfile,
+)
+from .transcript_builder import TranscriptBuilder
 
 logger = logging.getLogger(__name__)
 config = ChatConfig()
@@ -131,11 +129,6 @@ _CIRCUIT_BREAKER_ERROR_MSG = (
     "Try breaking your request into smaller parts."
 )
 
-# Idle timeout: abort the stream if no meaningful SDK message (only heartbeats)
-# arrives for this many seconds. This catches hung tool calls (e.g. WebSearch
-# hanging on a search provider that never responds).
-_IDLE_TIMEOUT_SECONDS = 10 * 60  # 10 minutes
-
 # Patterns that indicate the prompt/request exceeds the model's context limit.
 # Matched case-insensitively against the full exception chain.
 _PROMPT_TOO_LONG_PATTERNS: tuple[str, ...] = (
@@ -1278,8 +1271,6 @@ async def _run_stream_attempt(
             await client.query(state.query_message, session_id=ctx.session_id)
             state.transcript_builder.append_user(content=ctx.current_message)
 
-        _last_real_msg_time = time.monotonic()
-
         async for sdk_msg in _iter_sdk_messages(client):
             # Heartbeat sentinel — refresh lock and keep SSE alive
             if sdk_msg is None:
@@ -1287,34 +1278,8 @@ async def _run_stream_attempt(
                 for ev in ctx.compaction.emit_start_if_ready():
                     yield ev
                 yield StreamHeartbeat()
-
-                # Idle timeout: if no real SDK message for too long, a tool
-                # call is likely hung (e.g. WebSearch provider not responding).
-                idle_seconds = time.monotonic() - _last_real_msg_time
-                if idle_seconds >= _IDLE_TIMEOUT_SECONDS:
-                    logger.error(
-                        "%s Idle timeout after %.0fs with no SDK message — "
-                        "aborting stream (likely hung tool call)",
-                        ctx.log_prefix,
-                        idle_seconds,
-                    )
-                    stream_error_msg = (
-                        "A tool call appears to be stuck "
-                        "(no response for 10 minutes). "
-                        "Please try again."
-                    )
-                    stream_error_code = "idle_timeout"
-                    _append_error_marker(ctx.session, stream_error_msg, retryable=True)
-                    yield StreamError(
-                        errorText=stream_error_msg,
-                        code=stream_error_code,
-                    )
-                    ended_with_stream_error = True
-                    break
                 continue
 
-            _last_real_msg_time = time.monotonic()
-
             logger.info(
                 "%s Received: %s %s (unresolved=%d, current=%d, resolved=%d)",
                 ctx.log_prefix,
@@ -1563,21 +1528,9 @@ async def _run_stream_attempt(
             # --- Intermediate persistence ---
             # Flush session messages to DB periodically so page reloads
             # show progress during long-running turns.
-            #
-            # IMPORTANT: Skip the flush while tool calls are pending
-            # (tool_calls set on assistant but results not yet received).
-            # The DB save is append-only (uses start_sequence), so if we
-            # flush the assistant message before tool_calls are set on it
-            # (text and tool_use arrive as separate SDK events), the
-            # tool_calls update is lost — the next flush starts past it.
             _msgs_since_flush += 1
             now = time.monotonic()
-            has_pending_tools = (
-                acc.has_appended_assistant
-                and acc.accumulated_tool_calls
-                and not acc.has_tool_results
-            )
-            if not has_pending_tools and (
+            if (
                 _msgs_since_flush >= _FLUSH_MESSAGE_THRESHOLD
                 or (now - _last_flush_time) >= _FLUSH_INTERVAL_SECONDS
             ):
@@ -1677,7 +1630,6 @@ async def stream_chat_completion_sdk(
     session: ChatSession | None = None,
     file_ids: list[str] | None = None,
     permissions: "CopilotPermissions | None" = None,
-    mode: CopilotMode | None = None,
     **_kwargs: Any,
 ) -> AsyncIterator[StreamBaseResponse]:
     """Stream chat completion using Claude Agent SDK.
@@ -1686,10 +1638,7 @@ async def stream_chat_completion_sdk(
         file_ids: Optional workspace file IDs attached to the user's message.
             Images are embedded as vision content blocks; other files are
             saved to the SDK working directory for the Read tool.
-        mode: Accepted for signature compatibility with the baseline path.
-            The SDK path does not currently branch on this value.
     """
-    _ = mode  # SDK path ignores the requested mode.
 
     if session is None:
         session = await get_chat_session(session_id, user_id)
@@ -1720,12 +1669,19 @@ async def stream_chat_completion_sdk(
         )
         session.messages.pop()
 
-    if maybe_append_user_message(session, message, is_user_message):
+    # Append the new message to the session if it's not already there
+    new_message_role = "user" if is_user_message else "assistant"
+    if message and (
+        len(session.messages) == 0
+        or not (
+            session.messages[-1].role == new_message_role
+            and session.messages[-1].content == message
+        )
+    ):
+        session.messages.append(ChatMessage(role=new_message_role, content=message))
         if is_user_message:
             track_user_message(
-                user_id=user_id,
-                session_id=session_id,
-                message_length=len(message or ""),
+                user_id=user_id, session_id=session_id, message_length=len(message)
             )
 
     # Structured log prefix: [SDK][<session>][T<turn>]
@@ -1990,20 +1946,15 @@ async def stream_chat_completion_sdk(
         # langsmith tracing integration attaches them to every span.  This
         # is what Langfuse (or any OTEL backend) maps to its native
         # user/session fields.
-        _user_tier = await get_user_tier(user_id) if user_id else None
-        _otel_metadata: dict[str, str] = {
-            "resume": str(use_resume),
-            "conversation_turn": str(turn),
-        }
-        if _user_tier:
-            _otel_metadata["subscription_tier"] = _user_tier.value
-
         _otel_ctx = propagate_attributes(
             user_id=user_id,
             session_id=session_id,
             trace_name="copilot-sdk",
             tags=["sdk"],
-            metadata=_otel_metadata,
+            metadata={
+                "resume": str(use_resume),
+                "conversation_turn": str(turn),
+            },
         )
         _otel_ctx.__enter__()
 

autogpt_platform/backend/backend/copilot/sdk/service_test.py

@@ -10,6 +10,7 @@ import pytest
 
 from .service import (
     _is_sdk_disconnect_error,
+    _normalize_model_name,
     _prepare_file_attachments,
     _resolve_sdk_model,
     _safe_close_sdk_client,
@@ -405,6 +406,49 @@ def _clean_config_env(monkeypatch: pytest.MonkeyPatch) -> None:
         monkeypatch.delenv(var, raising=False)
 
 
+class TestNormalizeModelName:
+    """Tests for _normalize_model_name — shared provider-aware normalization."""
+
+    def test_strips_provider_prefix(self, monkeypatch, _clean_config_env):
+        from backend.copilot import config as cfg_mod
+
+        cfg = cfg_mod.ChatConfig(
+            use_openrouter=False,
+            api_key=None,
+            base_url=None,
+            use_claude_code_subscription=False,
+        )
+        monkeypatch.setattr("backend.copilot.sdk.service.config", cfg)
+        assert _normalize_model_name("anthropic/claude-opus-4.6") == "claude-opus-4-6"
+
+    def test_dots_preserved_for_openrouter(self, monkeypatch, _clean_config_env):
+        from backend.copilot import config as cfg_mod
+
+        cfg = cfg_mod.ChatConfig(
+            use_openrouter=True,
+            api_key="or-key",
+            base_url="https://openrouter.ai/api/v1",
+            use_claude_code_subscription=False,
+        )
+        monkeypatch.setattr("backend.copilot.sdk.service.config", cfg)
+        assert _normalize_model_name("anthropic/claude-opus-4.6") == "claude-opus-4.6"
+
+    def test_no_prefix_no_dots(self, monkeypatch, _clean_config_env):
+        from backend.copilot import config as cfg_mod
+
+        cfg = cfg_mod.ChatConfig(
+            use_openrouter=False,
+            api_key=None,
+            base_url=None,
+            use_claude_code_subscription=False,
+        )
+        monkeypatch.setattr("backend.copilot.sdk.service.config", cfg)
+        assert (
+            _normalize_model_name("claude-sonnet-4-20250514")
+            == "claude-sonnet-4-20250514"
+        )
+
+
 class TestResolveSdkModel:
     """Tests for _resolve_sdk_model — model ID resolution for the SDK CLI."""
 

autogpt_platform/backend/backend/copilot/sdk/thinking_blocks_test.py

@@ -27,19 +27,20 @@ from backend.copilot.response_model import (
     StreamTextDelta,
     StreamTextStart,
 )
-from backend.copilot.transcript import (
-    _find_last_assistant_entry,
-    _flatten_assistant_content,
-    _messages_to_transcript,
-    _rechain_tail,
-    _transcript_to_messages,
-)
 from backend.util import json
 
 from .conftest import build_structured_transcript
 from .response_adapter import SDKResponseAdapter
 from .service import _format_sdk_content_blocks
-from .transcript import compact_transcript, validate_transcript
+from .transcript import (
+    _find_last_assistant_entry,
+    _flatten_assistant_content,
+    _messages_to_transcript,
+    _rechain_tail,
+    _transcript_to_messages,
+    compact_transcript,
+    validate_transcript,
+)
 
 # ---------------------------------------------------------------------------
 # Fixtures: realistic thinking block content

autogpt_platform/backend/backend/copilot/sdk/transcript.py

@@ -1,25 +1,33 @@
-"""Re-export public API from shared ``backend.copilot.transcript``.
+"""Re-export from shared ``backend.copilot.transcript`` for backward compat.
 
 The canonical implementation now lives at ``backend.copilot.transcript``
 so both the SDK and baseline paths can import without cross-package
-dependencies.  Public symbols are re-exported here so existing ``from
+dependencies.  All symbols are re-exported here so existing ``from
 .transcript import ...`` statements within the ``sdk`` package continue
 to work without modification.
 """
 
 from backend.copilot.transcript import (
+    _MAX_PROJECT_DIRS_TO_SWEEP,
+    _STALE_PROJECT_DIR_SECONDS,
     COMPACT_MSG_ID_PREFIX,
     ENTRY_TYPE_MESSAGE,
     STOP_REASON_END_TURN,
     STRIPPABLE_TYPES,
     TRANSCRIPT_STORAGE_PREFIX,
     TranscriptDownload,
+    _find_last_assistant_entry,
+    _flatten_assistant_content,
+    _flatten_tool_result_content,
+    _messages_to_transcript,
+    _rechain_tail,
+    _run_compression,
+    _transcript_to_messages,
     cleanup_stale_project_dirs,
     compact_transcript,
     delete_transcript,
     download_transcript,
     read_compacted_entries,
-    strip_for_upload,
     strip_progress_entries,
     strip_stale_thinking_blocks,
     upload_transcript,
@@ -34,12 +42,20 @@ __all__ = [
     "STRIPPABLE_TYPES",
     "TRANSCRIPT_STORAGE_PREFIX",
     "TranscriptDownload",
+    "_MAX_PROJECT_DIRS_TO_SWEEP",
+    "_STALE_PROJECT_DIR_SECONDS",
+    "_find_last_assistant_entry",
+    "_flatten_assistant_content",
+    "_flatten_tool_result_content",
+    "_messages_to_transcript",
+    "_rechain_tail",
+    "_run_compression",
+    "_transcript_to_messages",
     "cleanup_stale_project_dirs",
     "compact_transcript",
     "delete_transcript",
     "download_transcript",
     "read_compacted_entries",
-    "strip_for_upload",
     "strip_progress_entries",
     "strip_stale_thinking_blocks",
     "upload_transcript",

autogpt_platform/backend/backend/copilot/sdk/transcript_test.py

@@ -850,7 +850,7 @@ class TestRunCompression:
     @pytest.mark.asyncio
     async def test_no_client_uses_truncation(self):
         """Path (a): ``get_openai_client()`` returns None → truncation only."""
-        from backend.copilot.transcript import _run_compression
+        from .transcript import _run_compression
 
         truncation_result = self._make_compress_result(
             True, [{"role": "user", "content": "truncated"}]
@@ -885,7 +885,7 @@ class TestRunCompression:
     @pytest.mark.asyncio
     async def test_llm_success_returns_llm_result(self):
         """Path (b): ``get_openai_client()`` returns a client → LLM compresses."""
-        from backend.copilot.transcript import _run_compression
+        from .transcript import _run_compression
 
         llm_result = self._make_compress_result(
             True, [{"role": "user", "content": "LLM summary"}]
@@ -916,7 +916,7 @@ class TestRunCompression:
     @pytest.mark.asyncio
     async def test_llm_failure_falls_back_to_truncation(self):
         """Path (c): LLM call raises → truncation fallback used instead."""
-        from backend.copilot.transcript import _run_compression
+        from .transcript import _run_compression
 
         truncation_result = self._make_compress_result(
             True, [{"role": "user", "content": "truncated fallback"}]
@@ -953,7 +953,7 @@ class TestRunCompression:
     @pytest.mark.asyncio
     async def test_llm_timeout_falls_back_to_truncation(self):
         """Path (d): LLM call exceeds timeout → truncation fallback used."""
-        from backend.copilot.transcript import _run_compression
+        from .transcript import _run_compression
 
         truncation_result = self._make_compress_result(
             True, [{"role": "user", "content": "truncated after timeout"}]
@@ -1007,7 +1007,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_removes_old_copilot_dirs(self, tmp_path, monkeypatch):
         """Directories matching copilot pattern older than threshold are removed."""
-        from backend.copilot.transcript import (
+        from backend.copilot.sdk.transcript import (
             _STALE_PROJECT_DIR_SECONDS,
             cleanup_stale_project_dirs,
         )
@@ -1039,7 +1039,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_ignores_non_copilot_dirs(self, tmp_path, monkeypatch):
         """Directories not matching copilot pattern are left alone."""
-        from backend.copilot.transcript import cleanup_stale_project_dirs
+        from backend.copilot.sdk.transcript import cleanup_stale_project_dirs
 
         projects_dir = tmp_path / "projects"
         projects_dir.mkdir()
@@ -1062,7 +1062,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_ttl_boundary_not_removed(self, tmp_path, monkeypatch):
         """A directory exactly at the TTL boundary should NOT be removed."""
-        from backend.copilot.transcript import (
+        from backend.copilot.sdk.transcript import (
             _STALE_PROJECT_DIR_SECONDS,
             cleanup_stale_project_dirs,
         )
@@ -1088,7 +1088,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_skips_non_directory_entries(self, tmp_path, monkeypatch):
         """Regular files matching the copilot pattern are not removed."""
-        from backend.copilot.transcript import (
+        from backend.copilot.sdk.transcript import (
             _STALE_PROJECT_DIR_SECONDS,
             cleanup_stale_project_dirs,
         )
@@ -1114,7 +1114,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_missing_base_dir_returns_zero(self, tmp_path, monkeypatch):
         """If the projects base directory doesn't exist, return 0 gracefully."""
-        from backend.copilot.transcript import cleanup_stale_project_dirs
+        from backend.copilot.sdk.transcript import cleanup_stale_project_dirs
 
         nonexistent = str(tmp_path / "does-not-exist" / "projects")
         monkeypatch.setattr(
@@ -1129,7 +1129,7 @@ class TestCleanupStaleProjectDirs:
         """When encoded_cwd is supplied only that directory is swept."""
         import time
 
-        from backend.copilot.transcript import (
+        from backend.copilot.sdk.transcript import (
             _STALE_PROJECT_DIR_SECONDS,
             cleanup_stale_project_dirs,
         )
@@ -1160,7 +1160,7 @@ class TestCleanupStaleProjectDirs:
 
     def test_scoped_fresh_dir_not_removed(self, tmp_path, monkeypatch):
         """Scoped sweep leaves a fresh directory alone."""
-        from backend.copilot.transcript import cleanup_stale_project_dirs
+        from backend.copilot.sdk.transcript import cleanup_stale_project_dirs
 
         projects_dir = tmp_path / "projects"
         projects_dir.mkdir()
@@ -1181,7 +1181,7 @@ class TestCleanupStaleProjectDirs:
         """Scoped sweep refuses to remove a non-copilot directory."""
         import time
 
-        from backend.copilot.transcript import (
+        from backend.copilot.sdk.transcript import (
             _STALE_PROJECT_DIR_SECONDS,
             cleanup_stale_project_dirs,
         )

autogpt_platform/backend/backend/copilot/token_tracking.py

@@ -4,12 +4,15 @@ Both the baseline (OpenRouter) and SDK (Anthropic) service layers need to:
   1. Append a ``Usage`` record to the session.
   2. Log the turn's token counts.
   3. Record weighted usage in Redis for rate-limiting.
+  4. Write a PlatformCostLog entry for admin cost tracking.
 
 This module extracts that common logic so both paths stay in sync.
 """
 
 import logging
 
+from backend.data.platform_cost import PlatformCostEntry, log_platform_cost_safe
+
 from .model import ChatSession, Usage
 from .rate_limit import record_token_usage
 
@@ -95,4 +98,47 @@ async def persist_and_record_usage(
         except Exception as usage_err:
             logger.warning(f"{log_prefix} Failed to record token usage: {usage_err}")
 
+    # Log to PlatformCostLog for admin cost dashboard
+    if user_id and total_tokens > 0:
+        cost_float = None
+        if cost_usd is not None:
+            try:
+                cost_float = float(cost_usd)
+            except (ValueError, TypeError):
+                pass
+
+        cost_microdollars = (
+            int(cost_float * 1_000_000) if cost_float is not None else None
+        )
+        session_id = session.session_id if session else None
+
+        if cost_float is not None:
+            tracking_type = "cost_usd"
+            tracking_amount = cost_float
+        else:
+            tracking_type = "tokens"
+            tracking_amount = total_tokens
+
+        await log_platform_cost_safe(
+            PlatformCostEntry(
+                user_id=user_id,
+                graph_exec_id=session_id,
+                block_id="copilot",
+                block_name=f"copilot:{log_prefix.strip(' []')}".rstrip(":"),
+                provider="open_router",
+                credential_id="copilot_system",
+                cost_microdollars=cost_microdollars,
+                input_tokens=prompt_tokens,
+                output_tokens=completion_tokens,
+                model=None,
+                metadata={
+                    "tracking_type": tracking_type,
+                    "tracking_amount": tracking_amount,
+                    "cache_read_tokens": cache_read_tokens,
+                    "cache_creation_tokens": cache_creation_tokens,
+                    "source": "copilot",
+                },
+            )
+        )
+
     return total_tokens

autogpt_platform/backend/backend/copilot/tools/agent_generator/fixer.py

@@ -33,12 +33,6 @@ _GET_CURRENT_DATE_BLOCK_ID = "b29c1b50-5d0e-4d9f-8f9d-1b0e6fcbf0b1"
 _GMAIL_SEND_BLOCK_ID = "6c27abc2-e51d-499e-a85f-5a0041ba94f0"
 _TEXT_REPLACE_BLOCK_ID = "7e7c87ab-3469-4bcc-9abe-67705091b713"
 
-# Default OrchestratorBlock model/mode — kept in sync with ChatConfig.model.
-# ChatConfig uses the OpenRouter format ("anthropic/claude-opus-4.6");
-# OrchestratorBlock uses the native Anthropic model name.
-ORCHESTRATOR_DEFAULT_MODEL = "claude-opus-4-6"
-ORCHESTRATOR_DEFAULT_EXECUTION_MODE = "extended_thinking"
-
 # Defaults applied to OrchestratorBlock nodes by the fixer.
 # execution_mode and model match the copilot's default (extended thinking
 # with Opus) so generated agents inherit the same reasoning capabilities.
@@ -48,8 +42,8 @@ _SDM_DEFAULTS: dict[str, int | bool | str] = {
     "conversation_compaction": True,
     "retry": 3,
     "multiple_tool_calls": False,
-    "execution_mode": ORCHESTRATOR_DEFAULT_EXECUTION_MODE,
-    "model": ORCHESTRATOR_DEFAULT_MODEL,
+    "execution_mode": "extended_thinking",
+    "model": "claude-opus-4-6",
 }
 
 
@@ -890,12 +884,6 @@ class AgentFixer:
             )
 
             if is_ai_block:
-                # Skip AI blocks that don't expose a "model" input property
-                # (some AI-category blocks have no model selector at all).
-                input_properties = block.get("inputSchema", {}).get("properties", {})
-                if "model" not in input_properties:
-                    continue
-
                 node_id = node.get("id")
                 input_default = node.get("input_default", {})
                 current_model = input_default.get("model")
@@ -904,7 +892,9 @@ class AgentFixer:
                 # Blocks with a block-specific enum on the model field (e.g.
                 # PerplexityBlock) use their own enum values; others use the
                 # generic set.
-                model_schema = input_properties.get("model", {})
+                model_schema = (
+                    block.get("inputSchema", {}).get("properties", {}).get("model", {})
+                )
                 block_model_enum = model_schema.get("enum")
 
                 if block_model_enum:
@@ -1765,12 +1755,6 @@ class AgentFixer:
         agent = self.fix_node_x_coordinates(agent, node_lookup=node_lookup)
         agent = self.fix_getcurrentdate_offset(agent)
 
-        # Apply OrchestratorBlock defaults BEFORE fix_ai_model_parameter so that
-        # the orchestrator-specific model (claude-opus-4-6) is set first and
-        # fix_ai_model_parameter sees it as a valid allowed model instead of
-        # overwriting it with the generic default (gpt-4o).
-        agent = self.fix_orchestrator_blocks(agent)
-
         # Apply fixes that require blocks information
         if blocks:
             agent = self.fix_invalid_nested_sink_links(
@@ -1788,6 +1772,9 @@ class AgentFixer:
         # Apply fixes for MCPToolBlock nodes
         agent = self.fix_mcp_tool_blocks(agent)
 
+        # Apply fixes for OrchestratorBlock nodes (agent-mode defaults)
+        agent = self.fix_orchestrator_blocks(agent)
+
         # Apply fixes for AgentExecutorBlock nodes (sub-agents)
         if library_agents:
             agent = self.fix_agent_executor_blocks(agent, library_agents)

autogpt_platform/backend/backend/copilot/tools/agent_generator/fixer_test.py

@@ -580,29 +580,6 @@ class TestFixAiModelParameter:
 
         assert result["nodes"][0]["input_default"]["model"] == "perplexity/sonar"
 
-    def test_ai_block_without_model_property_is_skipped(self):
-        """AI-category blocks that have no 'model' input property should not
-        have a model injected — they simply don't expose a model selector."""
-        fixer = AgentFixer()
-        block_id = generate_uuid()
-        node = _make_node(node_id="n1", block_id=block_id, input_default={})
-        agent = _make_agent(nodes=[node])
-
-        blocks = [
-            {
-                "id": block_id,
-                "name": "SomeAIBlock",
-                "categories": [{"category": "AI"}],
-                "inputSchema": {
-                    "properties": {"prompt": {"type": "string"}},
-                },
-            }
-        ]
-
-        result = fixer.fix_ai_model_parameter(agent, blocks)
-
-        assert "model" not in result["nodes"][0]["input_default"]
-
 
 class TestFixAgentExecutorBlocks:
     """Tests for fix_agent_executor_blocks."""

autogpt_platform/backend/backend/copilot/tools/get_agent_building_guide_test.py

@@ -1,15 +0,0 @@
-"""Tests for GetAgentBuildingGuideTool."""
-
-from backend.copilot.tools.get_agent_building_guide import _load_guide
-
-
-def test_load_guide_returns_string():
-    guide = _load_guide()
-    assert isinstance(guide, str)
-    assert len(guide) > 100
-
-
-def test_load_guide_caches():
-    guide1 = _load_guide()
-    guide2 = _load_guide()
-    assert guide1 is guide2

autogpt_platform/backend/backend/copilot/tools/test_dry_run.py

@@ -90,12 +90,11 @@ async def test_simulate_block_basic():
 
     with patch(
         "backend.executor.simulator.get_openai_client", return_value=mock_client
-    ) as mock_get_client:
+    ):
         outputs = []
         async for name, data in simulate_block(mock_block, {"query": "test"}):
             outputs.append((name, data))
 
-    mock_get_client.assert_called_once_with(prefer_openrouter=True)
     assert ("result", "simulated output") in outputs
     # Empty error pin should NOT be yielded — the simulator omits empty values
     assert ("error", "") not in outputs

autogpt_platform/backend/backend/copilot/tools/workspace_files.py

@@ -845,7 +845,6 @@ class WriteWorkspaceFileTool(BaseTool):
                 path=path,
                 mime_type=mime_type,
                 overwrite=overwrite,
-                metadata={"origin": "agent-created"},
             )
 
             # Build informative source label and message.

autogpt_platform/backend/backend/copilot/transcript.py

@@ -217,131 +217,6 @@ def strip_stale_thinking_blocks(content: str) -> str:
     return "\n".join(result_lines) + "\n"
 
 
-def strip_for_upload(content: str) -> str:
-    """Combined single-parse strip of progress entries and stale thinking blocks.
-
-    Equivalent to ``strip_stale_thinking_blocks(strip_progress_entries(content))``
-    but parses the JSONL only once, avoiding redundant ``split`` + ``json.loads``
-    passes on every upload.
-    """
-    lines = content.strip().split("\n")
-    if not lines:
-        return content
-
-    parsed: list[tuple[str, dict | None]] = []
-    for line in lines:
-        parsed.append((line, json.loads(line, fallback=None)))
-
-    # --- Phase 1: progress stripping (reparent children) ---
-    stripped_uuids: set[str] = set()
-    uuid_to_parent: dict[str, str] = {}
-
-    for _line, entry in parsed:
-        if not isinstance(entry, dict):
-            continue
-        uid = entry.get("uuid", "")
-        parent = entry.get("parentUuid", "")
-        if uid:
-            uuid_to_parent[uid] = parent
-        if (
-            entry.get("type", "") in STRIPPABLE_TYPES
-            and uid
-            and not entry.get("isCompactSummary")
-        ):
-            stripped_uuids.add(uid)
-
-    reparented: set[str] = set()
-    for _line, entry in parsed:
-        if not isinstance(entry, dict):
-            continue
-        parent = entry.get("parentUuid", "")
-        original_parent = parent
-        seen_parents: set[str] = set()
-        while parent in stripped_uuids and parent not in seen_parents:
-            seen_parents.add(parent)
-            parent = uuid_to_parent.get(parent, "")
-        if parent != original_parent:
-            entry["parentUuid"] = parent
-            uid = entry.get("uuid", "")
-            if uid:
-                reparented.add(uid)
-
-    # --- Phase 2: identify last assistant for thinking-block stripping ---
-    last_asst_msg_id: str | None = None
-    last_asst_idx: int | None = None
-    for i in range(len(parsed) - 1, -1, -1):
-        _line, entry = parsed[i]
-        if not isinstance(entry, dict):
-            continue
-        if entry.get("type", "") in STRIPPABLE_TYPES and not entry.get(
-            "isCompactSummary"
-        ):
-            continue
-        msg = entry.get("message", {})
-        if msg.get("role") == "assistant":
-            last_asst_msg_id = msg.get("id")
-            last_asst_idx = i
-            break
-
-    # --- Phase 3: single output pass ---
-    result_lines: list[str] = []
-    thinking_stripped = 0
-    for i, (line, entry) in enumerate(parsed):
-        if not isinstance(entry, dict):
-            result_lines.append(line)
-            continue
-
-        # Drop progress/metadata entries
-        if entry.get("type", "") in STRIPPABLE_TYPES and not entry.get(
-            "isCompactSummary"
-        ):
-            continue
-
-        needs_reserialize = False
-        uid = entry.get("uuid", "")
-
-        # Reparented entries need re-serialization
-        if uid in reparented:
-            needs_reserialize = True
-
-        # Strip stale thinking blocks from non-last assistant entries
-        if last_asst_idx is not None:
-            msg = entry.get("message", {})
-            is_last_turn = (
-                last_asst_msg_id is not None and msg.get("id") == last_asst_msg_id
-            ) or (last_asst_msg_id is None and i == last_asst_idx)
-            if (
-                msg.get("role") == "assistant"
-                and not is_last_turn
-                and isinstance(msg.get("content"), list)
-            ):
-                content_blocks = msg["content"]
-                filtered = [
-                    b
-                    for b in content_blocks
-                    if not (
-                        isinstance(b, dict) and b.get("type") in _THINKING_BLOCK_TYPES
-                    )
-                ]
-                if len(filtered) < len(content_blocks):
-                    thinking_stripped += len(content_blocks) - len(filtered)
-                    entry = {**entry, "message": {**msg, "content": filtered}}
-                    needs_reserialize = True
-
-        if needs_reserialize:
-            result_lines.append(json.dumps(entry, separators=(",", ":")))
-        else:
-            result_lines.append(line)
-
-    if thinking_stripped:
-        logger.info(
-            "[Transcript] Stripped %d stale thinking block(s) from non-last entries",
-            thinking_stripped,
-        )
-
-    return "\n".join(result_lines) + "\n"
-
-
 # ---------------------------------------------------------------------------
 # Local file I/O (write temp file for --resume)
 # ---------------------------------------------------------------------------
@@ -658,7 +533,6 @@ async def upload_transcript(
     content: str,
     message_count: int = 0,
     log_prefix: str = "[Transcript]",
-    skip_strip: bool = False,
 ) -> None:
     """Strip progress entries and stale thinking blocks, then upload transcript.
 
@@ -671,18 +545,14 @@ async def upload_transcript(
     Args:
         content: Complete JSONL transcript (from TranscriptBuilder).
         message_count: ``len(session.messages)`` at upload time.
-        skip_strip: When ``True``, skip the strip + re-validate pass.
-            Safe for builder-generated content (baseline path) which
-            never emits progress entries or stale thinking blocks.
     """
-    if skip_strip:
-        # Caller guarantees the content is already clean and valid.
-        stripped = content
-    else:
-        # Strip metadata entries and stale thinking blocks in a single parse.
-        # SDK-built transcripts may have progress entries; strip for safety.
-        stripped = strip_for_upload(content)
-    if not skip_strip and not validate_transcript(stripped):
+    # Strip metadata entries (progress, file-history-snapshot, etc.)
+    # Note: SDK-built transcripts shouldn't have these, but strip for safety.
+    # Also strip stale thinking blocks from non-last assistant entries to
+    # prevent token bloat when switching between SDK and baseline modes.
+    stripped = strip_progress_entries(content)
+    stripped = strip_stale_thinking_blocks(stripped)
+    if not validate_transcript(stripped):
         # Log entry types for debugging — helps identify why validation failed
         entry_types = [
             json.loads(line, fallback={"type": "INVALID_JSON"}).get("type", "?")
@@ -703,34 +573,27 @@ async def upload_transcript(
     storage = await get_workspace_storage()
     wid, fid, fname = _storage_path_parts(user_id, session_id)
     encoded = stripped.encode("utf-8")
-    meta = {"message_count": message_count, "uploaded_at": time.time()}
-    mwid, mfid, mfname = _meta_storage_path_parts(user_id, session_id)
-    meta_encoded = json.dumps(meta).encode("utf-8")
 
-    # Transcript + metadata are independent objects at different keys, so
-    # write them concurrently.  ``return_exceptions`` keeps a metadata
-    # failure from sinking the transcript write.
-    transcript_result, metadata_result = await asyncio.gather(
-        storage.store(
-            workspace_id=wid,
-            file_id=fid,
-            filename=fname,
-            content=encoded,
-        ),
-        storage.store(
+    await storage.store(
+        workspace_id=wid,
+        file_id=fid,
+        filename=fname,
+        content=encoded,
+    )
+
+    # Update metadata so message_count stays current.  The gap-fill logic
+    # in _build_query_message relies on it to avoid re-compressing messages.
+    try:
+        meta = {"message_count": message_count, "uploaded_at": time.time()}
+        mwid, mfid, mfname = _meta_storage_path_parts(user_id, session_id)
+        await storage.store(
             workspace_id=mwid,
             file_id=mfid,
             filename=mfname,
-            content=meta_encoded,
-        ),
-        return_exceptions=True,
-    )
-    if isinstance(transcript_result, BaseException):
-        raise transcript_result
-    if isinstance(metadata_result, BaseException):
-        # Metadata is best-effort — the gap-fill logic in
-        # _build_query_message tolerates a missing metadata file.
-        logger.warning("%s Failed to write metadata: %s", log_prefix, metadata_result)
+            content=json.dumps(meta).encode("utf-8"),
+        )
+    except Exception as e:
+        logger.warning("%s Failed to write metadata: %s", log_prefix, e)
 
     logger.info(
         "%s Uploaded %dB (stripped from %dB, msg_count=%d)",
@@ -750,44 +613,33 @@ async def download_transcript(
 
     Returns a ``TranscriptDownload`` with the JSONL content and the
     ``message_count`` watermark from the upload, or ``None`` if not found.
-
-    The content and metadata fetches run concurrently since they are
-    independent objects in the bucket.
     """
     storage = await get_workspace_storage()
     path = _build_storage_path(user_id, session_id, storage)
-    meta_path = _build_meta_storage_path(user_id, session_id, storage)
 
-    content_task = asyncio.create_task(storage.retrieve(path))
-    meta_task = asyncio.create_task(storage.retrieve(meta_path))
-    content_result, meta_result = await asyncio.gather(
-        content_task, meta_task, return_exceptions=True
-    )
-
-    if isinstance(content_result, FileNotFoundError):
+    try:
+        data = await storage.retrieve(path)
+        content = data.decode("utf-8")
+    except FileNotFoundError:
         logger.debug("%s No transcript in storage", log_prefix)
         return None
-    if isinstance(content_result, BaseException):
-        logger.warning(
-            "%s Failed to download transcript: %s", log_prefix, content_result
-        )
+    except Exception as e:
+        logger.warning("%s Failed to download transcript: %s", log_prefix, e)
         return None
 
-    content = content_result.decode("utf-8")
-
-    # Metadata is best-effort — old transcripts won't have it.
+    # Try to load metadata (best-effort — old transcripts won't have it)
     message_count = 0
     uploaded_at = 0.0
-    if isinstance(meta_result, FileNotFoundError):
-        pass  # No metadata — treat as unknown (msg_count=0 → always fill gap)
-    elif isinstance(meta_result, BaseException):
-        logger.debug(
-            "%s Failed to load transcript metadata: %s", log_prefix, meta_result
-        )
-    else:
-        meta = json.loads(meta_result.decode("utf-8"), fallback={})
+    try:
+        meta_path = _build_meta_storage_path(user_id, session_id, storage)
+        meta_data = await storage.retrieve(meta_path)
+        meta = json.loads(meta_data.decode("utf-8"), fallback={})
         message_count = meta.get("message_count", 0)
         uploaded_at = meta.get("uploaded_at", 0.0)
+    except FileNotFoundError:
+        pass  # No metadata — treat as unknown (msg_count=0 → always fill gap)
+    except Exception as e:
+        logger.debug("%s Failed to load transcript metadata: %s", log_prefix, e)
 
     logger.info(
         "%s Downloaded %dB (msg_count=%d)", log_prefix, len(content), message_count
@@ -829,7 +681,6 @@ async def delete_transcript(user_id: str, session_id: str) -> None:
 
 # JSONL protocol values used in transcript serialization.
 STOP_REASON_END_TURN = "end_turn"
-STOP_REASON_TOOL_USE = "tool_use"
 COMPACT_MSG_ID_PREFIX = "msg_compact_"
 ENTRY_TYPE_MESSAGE = "message"
 

autogpt_platform/backend/backend/copilot/transcript_builder.py

@@ -29,7 +29,7 @@ class TranscriptEntry(BaseModel):
 
     type: str
     uuid: str
-    parentUuid: str = ""
+    parentUuid: str | None
     isCompactSummary: bool | None = None
     message: dict[str, Any]
 
@@ -67,7 +67,7 @@ class TranscriptBuilder:
         return TranscriptEntry(
             type=entry_type,
             uuid=data.get("uuid") or str(uuid4()),
-            parentUuid=data.get("parentUuid") or "",
+            parentUuid=data.get("parentUuid"),
             isCompactSummary=data.get("isCompactSummary"),
             message=data.get("message", {}),
         )
@@ -118,7 +118,7 @@ class TranscriptBuilder:
             TranscriptEntry(
                 type="user",
                 uuid=msg_uuid,
-                parentUuid=self._last_uuid or "",
+                parentUuid=self._last_uuid,
                 message={"role": "user", "content": content},
             )
         )
@@ -158,7 +158,7 @@ class TranscriptBuilder:
             TranscriptEntry(
                 type="assistant",
                 uuid=msg_uuid,
-                parentUuid=self._last_uuid or "",
+                parentUuid=self._last_uuid,
                 message={
                     "role": "assistant",
                     "model": model,
@@ -233,8 +233,3 @@ class TranscriptBuilder:
     def is_empty(self) -> bool:
         """Whether this builder has any entries."""
         return len(self._entries) == 0
-
-    @property
-    def last_entry_type(self) -> str | None:
-        """Type of the last entry, or None if empty."""
-        return self._entries[-1].type if self._entries else None