mirror of
https://github.com/Significant-Gravitas/AutoGPT.git
synced 2026-04-08 03:00:28 -04:00
Compare commits
1 Commits
combined-p
...
master
| Author | SHA1 | Date | |
|---|---|---|---|
|
1750c833ee |
@@ -1 +0,0 @@
|
||||
../.claude/skills
|
||||
@@ -1,106 +0,0 @@
|
||||
---
|
||||
name: open-pr
|
||||
description: Open a pull request with proper PR template, test coverage, and review workflow. Guides agents through creating a PR that follows repo conventions, ensures existing behaviors aren't broken, covers new behaviors with tests, and handles review via bot when local testing isn't possible. TRIGGER when user asks to "open a PR", "create a PR", "make a PR", "submit a PR", "open pull request", "push and create PR", or any variation of opening/submitting a pull request.
|
||||
user-invocable: true
|
||||
args: "[base-branch] — optional target branch (defaults to dev)."
|
||||
metadata:
|
||||
author: autogpt-team
|
||||
version: "1.0.0"
|
||||
---
|
||||
|
||||
# Open a Pull Request
|
||||
|
||||
## Step 1: Pre-flight checks
|
||||
|
||||
Before opening the PR:
|
||||
|
||||
1. Ensure all changes are committed
|
||||
2. Ensure the branch is pushed to the remote (`git push -u origin <branch>`)
|
||||
3. Run linters/formatters across the whole repo (not just changed files) and commit any fixes
|
||||
|
||||
## Step 2: Test coverage
|
||||
|
||||
**This is critical.** Before opening the PR, verify:
|
||||
|
||||
### Existing behavior is not broken
|
||||
- Identify which modules/components your changes touch
|
||||
- Run the existing test suites for those areas
|
||||
- If tests fail, fix them before opening the PR — do not open a PR with known regressions
|
||||
|
||||
### New behavior has test coverage
|
||||
- Every new feature, endpoint, or behavior change needs tests
|
||||
- If you added a new block, add tests for that block
|
||||
- If you changed API behavior, add or update API tests
|
||||
- If you changed frontend behavior, verify it doesn't break existing flows
|
||||
|
||||
If you cannot run the full test suite locally, note which tests you ran and which you couldn't in the test plan.
|
||||
|
||||
## Step 3: Create the PR using the repo template
|
||||
|
||||
Read the canonical PR template at `.github/PULL_REQUEST_TEMPLATE.md` and use it **verbatim** as your PR body:
|
||||
|
||||
1. Read the template: `cat .github/PULL_REQUEST_TEMPLATE.md`
|
||||
2. Preserve the exact section titles and formatting, including:
|
||||
- `### Why / What / How`
|
||||
- `### Changes 🏗️`
|
||||
- `### Checklist 📋`
|
||||
3. Replace HTML comment prompts (`<!-- ... -->`) with actual content; do not leave them in
|
||||
4. **Do not pre-check boxes** — leave all checkboxes as `- [ ]` until each step is actually completed
|
||||
5. Do not alter the template structure, rename sections, or remove any checklist items
|
||||
|
||||
**PR title must use conventional commit format** (e.g., `feat(backend): add new block`, `fix(frontend): resolve routing bug`, `dx(skills): update PR workflow`). See CLAUDE.md for the full list of scopes.
|
||||
|
||||
Use `gh pr create` with the base branch (defaults to `dev` if no `[base-branch]` was provided). Use `--body-file` to avoid shell interpretation of backticks and special characters:
|
||||
|
||||
```bash
|
||||
BASE_BRANCH="${BASE_BRANCH:-dev}"
|
||||
PR_BODY=$(mktemp)
|
||||
cat > "$PR_BODY" << 'PREOF'
|
||||
<filled-in template from .github/PULL_REQUEST_TEMPLATE.md>
|
||||
PREOF
|
||||
gh pr create --base "$BASE_BRANCH" --title "<type>(scope): short description" --body-file "$PR_BODY"
|
||||
rm "$PR_BODY"
|
||||
```
|
||||
|
||||
## Step 4: Review workflow
|
||||
|
||||
### If you have a workspace that allows testing (docker, running backend, etc.)
|
||||
- Run `/pr-test` to do E2E manual testing of the PR using docker compose, agent-browser, and API calls. This is the most thorough way to validate your changes before review.
|
||||
- After testing, run `/pr-review` to self-review the PR for correctness, security, code quality, and testing gaps before requesting human review.
|
||||
|
||||
### If you do NOT have a workspace that allows testing
|
||||
This is common for agents running in worktrees without a full stack. In this case:
|
||||
|
||||
1. Run `/pr-review` locally to catch obvious issues before pushing
|
||||
2. **Comment `/review` on the PR** after creating it to trigger the review bot
|
||||
3. **Poll for the review** rather than blindly waiting — check for new review comments every 30 seconds using `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/reviews --paginate` and the GraphQL inline threads query. The bot typically responds within 30 minutes, but polling lets the agent react as soon as it arrives.
|
||||
4. Do NOT proceed or merge until the bot review comes back
|
||||
5. Address any issues the bot raises — use `/pr-address` which has a full polling loop with CI + comment tracking
|
||||
|
||||
```bash
|
||||
# After creating the PR:
|
||||
PR_NUMBER=$(gh pr view --json number -q .number)
|
||||
gh pr comment "$PR_NUMBER" --body "/review"
|
||||
# Then use /pr-address to poll for and address the review when it arrives
|
||||
```
|
||||
|
||||
## Step 5: Address review feedback
|
||||
|
||||
Once the review bot or human reviewers leave comments:
|
||||
- Run `/pr-address` to address review comments. It will loop until CI is green and all comments are resolved.
|
||||
- Do not merge without human approval.
|
||||
|
||||
## Related skills
|
||||
|
||||
| Skill | When to use |
|
||||
|---|---|
|
||||
| `/pr-test` | E2E testing with docker compose, agent-browser, API calls — use when you have a running workspace |
|
||||
| `/pr-review` | Review for correctness, security, code quality — use before requesting human review |
|
||||
| `/pr-address` | Address reviewer comments and loop until CI green — use after reviews come in |
|
||||
|
||||
## Step 6: Post-creation
|
||||
|
||||
After the PR is created and review is triggered:
|
||||
- Share the PR URL with the user
|
||||
- If waiting on the review bot, let the user know the expected wait time (~30 min)
|
||||
- Do not merge without human approval
|
||||
@@ -1,195 +0,0 @@
|
||||
---
|
||||
name: setup-repo
|
||||
description: Initialize a worktree-based repo layout for parallel development. Creates a main worktree, a reviews worktree for PR reviews, and N numbered work branches. Handles .env creation, dependency installation, and branchlet config. TRIGGER when user asks to set up the repo from scratch, initialize worktrees, bootstrap their dev environment, "setup repo", "setup worktrees", "initialize dev environment", "set up branches", or when a freshly cloned repo has no sibling worktrees.
|
||||
user-invocable: true
|
||||
args: "No arguments — interactive setup via prompts."
|
||||
metadata:
|
||||
author: autogpt-team
|
||||
version: "1.0.0"
|
||||
---
|
||||
|
||||
# Repository Setup
|
||||
|
||||
This skill sets up a worktree-based development layout from a freshly cloned repo. It creates:
|
||||
- A **main** worktree (the primary checkout)
|
||||
- A **reviews** worktree (for PR reviews)
|
||||
- **N work branches** (branch1..branchN) for parallel development
|
||||
|
||||
## Step 1: Identify the repo
|
||||
|
||||
Determine the repo root and parent directory:
|
||||
|
||||
```bash
|
||||
ROOT=$(git rev-parse --show-toplevel)
|
||||
REPO_NAME=$(basename "$ROOT")
|
||||
PARENT=$(dirname "$ROOT")
|
||||
```
|
||||
|
||||
Detect if the repo is already inside a worktree layout by counting sibling worktrees (not just checking the directory name, which could be anything):
|
||||
|
||||
```bash
|
||||
# Count worktrees that are siblings (live under $PARENT but aren't $ROOT itself)
|
||||
SIBLING_COUNT=$(git worktree list --porcelain 2>/dev/null | grep "^worktree " | grep -c "$PARENT/" || true)
|
||||
if [ "$SIBLING_COUNT" -gt 1 ]; then
|
||||
echo "INFO: Existing worktree layout detected at $PARENT ($SIBLING_COUNT worktrees)"
|
||||
# Use $ROOT as-is; skip renaming/restructuring
|
||||
else
|
||||
echo "INFO: Fresh clone detected, proceeding with setup"
|
||||
fi
|
||||
```
|
||||
|
||||
## Step 2: Ask the user questions
|
||||
|
||||
Use AskUserQuestion to gather setup preferences:
|
||||
|
||||
1. **How many parallel work branches do you need?** (Options: 4, 8, 16, or custom)
|
||||
- These become `branch1` through `branchN`
|
||||
2. **Which branch should be the base?** (Options: origin/master, origin/dev, or custom)
|
||||
- All work branches and reviews will start from this
|
||||
|
||||
## Step 3: Fetch and set up branches
|
||||
|
||||
```bash
|
||||
cd "$ROOT"
|
||||
git fetch origin
|
||||
|
||||
# Create the reviews branch from base (skip if already exists)
|
||||
if git show-ref --verify --quiet refs/heads/reviews; then
|
||||
echo "INFO: Branch 'reviews' already exists, skipping"
|
||||
else
|
||||
git branch reviews <base-branch>
|
||||
fi
|
||||
|
||||
# Create numbered work branches from base (skip if already exists)
|
||||
for i in $(seq 1 "$COUNT"); do
|
||||
if git show-ref --verify --quiet "refs/heads/branch$i"; then
|
||||
echo "INFO: Branch 'branch$i' already exists, skipping"
|
||||
else
|
||||
git branch "branch$i" <base-branch>
|
||||
fi
|
||||
done
|
||||
```
|
||||
|
||||
## Step 4: Create worktrees
|
||||
|
||||
Create worktrees as siblings to the main checkout:
|
||||
|
||||
```bash
|
||||
if [ -d "$PARENT/reviews" ]; then
|
||||
echo "INFO: Worktree '$PARENT/reviews' already exists, skipping"
|
||||
else
|
||||
git worktree add "$PARENT/reviews" reviews
|
||||
fi
|
||||
|
||||
for i in $(seq 1 "$COUNT"); do
|
||||
if [ -d "$PARENT/branch$i" ]; then
|
||||
echo "INFO: Worktree '$PARENT/branch$i' already exists, skipping"
|
||||
else
|
||||
git worktree add "$PARENT/branch$i" "branch$i"
|
||||
fi
|
||||
done
|
||||
```
|
||||
|
||||
## Step 5: Set up environment files
|
||||
|
||||
**Do NOT assume .env files exist.** For each worktree (including main if needed):
|
||||
|
||||
1. Check if `.env` exists in the source worktree for each path
|
||||
2. If `.env` exists, copy it
|
||||
3. If only `.env.default` or `.env.example` exists, copy that as `.env`
|
||||
4. If neither exists, warn the user and list which env files are missing
|
||||
|
||||
Env file locations to check (same as the `/worktree` skill — keep these in sync):
|
||||
- `autogpt_platform/.env`
|
||||
- `autogpt_platform/backend/.env`
|
||||
- `autogpt_platform/frontend/.env`
|
||||
|
||||
> **Note:** This env copying logic intentionally mirrors the `/worktree` skill's approach. If you update the path list or fallback logic here, update `/worktree` as well.
|
||||
|
||||
```bash
|
||||
SOURCE="$ROOT"
|
||||
WORKTREES="reviews"
|
||||
for i in $(seq 1 "$COUNT"); do WORKTREES="$WORKTREES branch$i"; done
|
||||
|
||||
FOUND_ANY_ENV=0
|
||||
for wt in $WORKTREES; do
|
||||
TARGET="$PARENT/$wt"
|
||||
for envpath in autogpt_platform autogpt_platform/backend autogpt_platform/frontend; do
|
||||
if [ -f "$SOURCE/$envpath/.env" ]; then
|
||||
FOUND_ANY_ENV=1
|
||||
cp "$SOURCE/$envpath/.env" "$TARGET/$envpath/.env"
|
||||
elif [ -f "$SOURCE/$envpath/.env.default" ]; then
|
||||
FOUND_ANY_ENV=1
|
||||
cp "$SOURCE/$envpath/.env.default" "$TARGET/$envpath/.env"
|
||||
echo "NOTE: $wt/$envpath/.env was created from .env.default — you may need to edit it"
|
||||
elif [ -f "$SOURCE/$envpath/.env.example" ]; then
|
||||
FOUND_ANY_ENV=1
|
||||
cp "$SOURCE/$envpath/.env.example" "$TARGET/$envpath/.env"
|
||||
echo "NOTE: $wt/$envpath/.env was created from .env.example — you may need to edit it"
|
||||
else
|
||||
echo "WARNING: No .env, .env.default, or .env.example found at $SOURCE/$envpath/"
|
||||
fi
|
||||
done
|
||||
done
|
||||
|
||||
if [ "$FOUND_ANY_ENV" -eq 0 ]; then
|
||||
echo "WARNING: No environment files or templates were found in the source worktree."
|
||||
# Use AskUserQuestion to confirm: "Continue setup without env files?"
|
||||
# If the user declines, stop here and let them set up .env files first.
|
||||
fi
|
||||
```
|
||||
|
||||
## Step 6: Copy branchlet config
|
||||
|
||||
Copy `.branchlet.json` from main to each worktree so branchlet can manage sub-worktrees:
|
||||
|
||||
```bash
|
||||
if [ -f "$ROOT/.branchlet.json" ]; then
|
||||
for wt in $WORKTREES; do
|
||||
cp "$ROOT/.branchlet.json" "$PARENT/$wt/.branchlet.json"
|
||||
done
|
||||
fi
|
||||
```
|
||||
|
||||
## Step 7: Install dependencies
|
||||
|
||||
Install deps in all worktrees. Run these sequentially per worktree:
|
||||
|
||||
```bash
|
||||
for wt in $WORKTREES; do
|
||||
TARGET="$PARENT/$wt"
|
||||
echo "=== Installing deps for $wt ==="
|
||||
(cd "$TARGET/autogpt_platform/autogpt_libs" && poetry install) &&
|
||||
(cd "$TARGET/autogpt_platform/backend" && poetry install && poetry run prisma generate) &&
|
||||
(cd "$TARGET/autogpt_platform/frontend" && pnpm install) &&
|
||||
echo "=== Done: $wt ===" ||
|
||||
echo "=== FAILED: $wt ==="
|
||||
done
|
||||
```
|
||||
|
||||
This is slow. Run in background if possible and notify when complete.
|
||||
|
||||
## Step 8: Verify and report
|
||||
|
||||
After setup, verify and report to the user:
|
||||
|
||||
```bash
|
||||
git worktree list
|
||||
```
|
||||
|
||||
Summarize:
|
||||
- Number of worktrees created
|
||||
- Which env files were copied vs created from defaults vs missing
|
||||
- Any warnings or errors encountered
|
||||
|
||||
## Final directory layout
|
||||
|
||||
```
|
||||
parent/
|
||||
main/ # Primary checkout (already exists)
|
||||
reviews/ # PR review worktree
|
||||
branch1/ # Work branch 1
|
||||
branch2/ # Work branch 2
|
||||
...
|
||||
branchN/ # Work branch N
|
||||
```
|
||||
@@ -1,6 +1,6 @@
|
||||
# AutoGPT Platform Contribution Guide
|
||||
|
||||
This guide provides context for coding agents when updating the **autogpt_platform** folder.
|
||||
This guide provides context for Codex when updating the **autogpt_platform** folder.
|
||||
|
||||
## Directory overview
|
||||
|
||||
|
||||
@@ -83,13 +83,13 @@ The AutoGPT frontend is where users interact with our powerful AI automation pla
|
||||
|
||||
**Agent Builder:** For those who want to customize, our intuitive, low-code interface allows you to design and configure your own AI agents.
|
||||
|
||||
**Workflow Management:** Build, modify, and optimize your automation workflows with ease. You build your agent by connecting blocks, where each block performs a single action.
|
||||
**Workflow Management:** Build, modify, and optimize your automation workflows with ease. You build your agent by connecting blocks, where each block performs a single action.
|
||||
|
||||
**Deployment Controls:** Manage the lifecycle of your agents, from testing to production.
|
||||
|
||||
**Ready-to-Use Agents:** Don't want to build? Simply select from our library of pre-configured agents and put them to work immediately.
|
||||
|
||||
**Agent Interaction:** Whether you've built your own or are using pre-configured agents, easily run and interact with them through our user-friendly interface.
|
||||
**Agent Interaction:** Whether you've built your own or are using pre-configured agents, easily run and interact with them through our user-friendly interface.
|
||||
|
||||
**Monitoring and Analytics:** Keep track of your agents' performance and gain insights to continually improve your automation processes.
|
||||
|
||||
|
||||
@@ -1,120 +0,0 @@
|
||||
# AutoGPT Platform
|
||||
|
||||
This file provides guidance to coding agents when working with code in this repository.
|
||||
|
||||
## Repository Overview
|
||||
|
||||
AutoGPT Platform is a monorepo containing:
|
||||
|
||||
- **Backend** (`backend`): Python FastAPI server with async support
|
||||
- **Frontend** (`frontend`): Next.js React application
|
||||
- **Shared Libraries** (`autogpt_libs`): Common Python utilities
|
||||
|
||||
## Component Documentation
|
||||
|
||||
- **Backend**: See @backend/AGENTS.md for backend-specific commands, architecture, and development tasks
|
||||
- **Frontend**: See @frontend/AGENTS.md for frontend-specific commands, architecture, and development patterns
|
||||
|
||||
## Key Concepts
|
||||
|
||||
1. **Agent Graphs**: Workflow definitions stored as JSON, executed by the backend
|
||||
2. **Blocks**: Reusable components in `backend/backend/blocks/` that perform specific tasks
|
||||
3. **Integrations**: OAuth and API connections stored per user
|
||||
4. **Store**: Marketplace for sharing agent templates
|
||||
5. **Virus Scanning**: ClamAV integration for file upload security
|
||||
|
||||
### Environment Configuration
|
||||
|
||||
#### Configuration Files
|
||||
|
||||
- **Backend**: `backend/.env.default` (defaults) → `backend/.env` (user overrides)
|
||||
- **Frontend**: `frontend/.env.default` (defaults) → `frontend/.env` (user overrides)
|
||||
- **Platform**: `.env.default` (Supabase/shared defaults) → `.env` (user overrides)
|
||||
|
||||
#### Docker Environment Loading Order
|
||||
|
||||
1. `.env.default` files provide base configuration (tracked in git)
|
||||
2. `.env` files provide user-specific overrides (gitignored)
|
||||
3. Docker Compose `environment:` sections provide service-specific overrides
|
||||
4. Shell environment variables have highest precedence
|
||||
|
||||
#### Key Points
|
||||
|
||||
- All services use hardcoded defaults in docker-compose files (no `${VARIABLE}` substitutions)
|
||||
- The `env_file` directive loads variables INTO containers at runtime
|
||||
- Backend/Frontend services use YAML anchors for consistent configuration
|
||||
- Supabase services (`db/docker/docker-compose.yml`) follow the same pattern
|
||||
|
||||
### Branching Strategy
|
||||
|
||||
- **`dev`** is the main development branch. All PRs should target `dev`.
|
||||
- **`master`** is the production branch. Only used for production releases.
|
||||
|
||||
### Creating Pull Requests
|
||||
|
||||
- Create the PR against the `dev` branch of the repository.
|
||||
- **Split PRs by concern** — each PR should have a single clear purpose. For example, "usage tracking" and "credit charging" should be separate PRs even if related. Combining multiple concerns makes it harder for reviewers to understand what belongs to what.
|
||||
- Ensure the branch name is descriptive (e.g., `feature/add-new-block`)
|
||||
- Use conventional commit messages (see below)
|
||||
- **Structure the PR description with Why / What / How** — Why: the motivation (what problem it solves, what's broken/missing without it); What: high-level summary of changes; How: approach, key implementation details, or architecture decisions. Reviewers need all three to judge whether the approach fits the problem.
|
||||
- Fill out the .github/PULL_REQUEST_TEMPLATE.md template as the PR description
|
||||
- Always use `--body-file` to pass PR body — avoids shell interpretation of backticks and special characters:
|
||||
```bash
|
||||
PR_BODY=$(mktemp)
|
||||
cat > "$PR_BODY" << 'PREOF'
|
||||
## Summary
|
||||
- use `backticks` freely here
|
||||
PREOF
|
||||
gh pr create --title "..." --body-file "$PR_BODY" --base dev
|
||||
rm "$PR_BODY"
|
||||
```
|
||||
- Run the github pre-commit hooks to ensure code quality.
|
||||
|
||||
### Test-Driven Development (TDD)
|
||||
|
||||
When fixing a bug or adding a feature, follow a test-first approach:
|
||||
|
||||
1. **Write a failing test first** — create a test that reproduces the bug or validates the new behavior, marked with `@pytest.mark.xfail` (backend) or `.fixme` (Playwright). Run it to confirm it fails for the right reason.
|
||||
2. **Implement the fix/feature** — write the minimal code to make the test pass.
|
||||
3. **Remove the xfail marker** — once the test passes, remove the `xfail`/`.fixme` annotation and run the full test suite to confirm nothing else broke.
|
||||
|
||||
This ensures every change is covered by a test and that the test actually validates the intended behavior.
|
||||
|
||||
### Reviewing/Revising Pull Requests
|
||||
|
||||
Use `/pr-review` to review a PR or `/pr-address` to address comments.
|
||||
|
||||
When fetching comments manually:
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/reviews --paginate` — top-level reviews
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/comments --paginate` — inline review comments (always paginate to avoid missing comments beyond page 1)
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/issues/{N}/comments` — PR conversation comments
|
||||
|
||||
### Conventional Commits
|
||||
|
||||
Use this format for commit messages and Pull Request titles:
|
||||
|
||||
**Conventional Commit Types:**
|
||||
|
||||
- `feat`: Introduces a new feature to the codebase
|
||||
- `fix`: Patches a bug in the codebase
|
||||
- `refactor`: Code change that neither fixes a bug nor adds a feature; also applies to removing features
|
||||
- `ci`: Changes to CI configuration
|
||||
- `docs`: Documentation-only changes
|
||||
- `dx`: Improvements to the developer experience
|
||||
|
||||
**Recommended Base Scopes:**
|
||||
|
||||
- `platform`: Changes affecting both frontend and backend
|
||||
- `frontend`
|
||||
- `backend`
|
||||
- `infra`
|
||||
- `blocks`: Modifications/additions of individual blocks
|
||||
|
||||
**Subscope Examples:**
|
||||
|
||||
- `backend/executor`
|
||||
- `backend/db`
|
||||
- `frontend/builder` (includes changes to the block UI component)
|
||||
- `infra/prod`
|
||||
|
||||
Use these scopes and subscopes for clarity and consistency in commit messages.
|
||||
@@ -1 +1,120 @@
|
||||
@AGENTS.md
|
||||
# CLAUDE.md
|
||||
|
||||
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
|
||||
|
||||
## Repository Overview
|
||||
|
||||
AutoGPT Platform is a monorepo containing:
|
||||
|
||||
- **Backend** (`backend`): Python FastAPI server with async support
|
||||
- **Frontend** (`frontend`): Next.js React application
|
||||
- **Shared Libraries** (`autogpt_libs`): Common Python utilities
|
||||
|
||||
## Component Documentation
|
||||
|
||||
- **Backend**: See @backend/CLAUDE.md for backend-specific commands, architecture, and development tasks
|
||||
- **Frontend**: See @frontend/CLAUDE.md for frontend-specific commands, architecture, and development patterns
|
||||
|
||||
## Key Concepts
|
||||
|
||||
1. **Agent Graphs**: Workflow definitions stored as JSON, executed by the backend
|
||||
2. **Blocks**: Reusable components in `backend/backend/blocks/` that perform specific tasks
|
||||
3. **Integrations**: OAuth and API connections stored per user
|
||||
4. **Store**: Marketplace for sharing agent templates
|
||||
5. **Virus Scanning**: ClamAV integration for file upload security
|
||||
|
||||
### Environment Configuration
|
||||
|
||||
#### Configuration Files
|
||||
|
||||
- **Backend**: `backend/.env.default` (defaults) → `backend/.env` (user overrides)
|
||||
- **Frontend**: `frontend/.env.default` (defaults) → `frontend/.env` (user overrides)
|
||||
- **Platform**: `.env.default` (Supabase/shared defaults) → `.env` (user overrides)
|
||||
|
||||
#### Docker Environment Loading Order
|
||||
|
||||
1. `.env.default` files provide base configuration (tracked in git)
|
||||
2. `.env` files provide user-specific overrides (gitignored)
|
||||
3. Docker Compose `environment:` sections provide service-specific overrides
|
||||
4. Shell environment variables have highest precedence
|
||||
|
||||
#### Key Points
|
||||
|
||||
- All services use hardcoded defaults in docker-compose files (no `${VARIABLE}` substitutions)
|
||||
- The `env_file` directive loads variables INTO containers at runtime
|
||||
- Backend/Frontend services use YAML anchors for consistent configuration
|
||||
- Supabase services (`db/docker/docker-compose.yml`) follow the same pattern
|
||||
|
||||
### Branching Strategy
|
||||
|
||||
- **`dev`** is the main development branch. All PRs should target `dev`.
|
||||
- **`master`** is the production branch. Only used for production releases.
|
||||
|
||||
### Creating Pull Requests
|
||||
|
||||
- Create the PR against the `dev` branch of the repository.
|
||||
- **Split PRs by concern** — each PR should have a single clear purpose. For example, "usage tracking" and "credit charging" should be separate PRs even if related. Combining multiple concerns makes it harder for reviewers to understand what belongs to what.
|
||||
- Ensure the branch name is descriptive (e.g., `feature/add-new-block`)
|
||||
- Use conventional commit messages (see below)
|
||||
- **Structure the PR description with Why / What / How** — Why: the motivation (what problem it solves, what's broken/missing without it); What: high-level summary of changes; How: approach, key implementation details, or architecture decisions. Reviewers need all three to judge whether the approach fits the problem.
|
||||
- Fill out the .github/PULL_REQUEST_TEMPLATE.md template as the PR description
|
||||
- Always use `--body-file` to pass PR body — avoids shell interpretation of backticks and special characters:
|
||||
```bash
|
||||
PR_BODY=$(mktemp)
|
||||
cat > "$PR_BODY" << 'PREOF'
|
||||
## Summary
|
||||
- use `backticks` freely here
|
||||
PREOF
|
||||
gh pr create --title "..." --body-file "$PR_BODY" --base dev
|
||||
rm "$PR_BODY"
|
||||
```
|
||||
- Run the github pre-commit hooks to ensure code quality.
|
||||
|
||||
### Test-Driven Development (TDD)
|
||||
|
||||
When fixing a bug or adding a feature, follow a test-first approach:
|
||||
|
||||
1. **Write a failing test first** — create a test that reproduces the bug or validates the new behavior, marked with `@pytest.mark.xfail` (backend) or `.fixme` (Playwright). Run it to confirm it fails for the right reason.
|
||||
2. **Implement the fix/feature** — write the minimal code to make the test pass.
|
||||
3. **Remove the xfail marker** — once the test passes, remove the `xfail`/`.fixme` annotation and run the full test suite to confirm nothing else broke.
|
||||
|
||||
This ensures every change is covered by a test and that the test actually validates the intended behavior.
|
||||
|
||||
### Reviewing/Revising Pull Requests
|
||||
|
||||
Use `/pr-review` to review a PR or `/pr-address` to address comments.
|
||||
|
||||
When fetching comments manually:
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/reviews --paginate` — top-level reviews
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/comments --paginate` — inline review comments (always paginate to avoid missing comments beyond page 1)
|
||||
- `gh api repos/Significant-Gravitas/AutoGPT/issues/{N}/comments` — PR conversation comments
|
||||
|
||||
### Conventional Commits
|
||||
|
||||
Use this format for commit messages and Pull Request titles:
|
||||
|
||||
**Conventional Commit Types:**
|
||||
|
||||
- `feat`: Introduces a new feature to the codebase
|
||||
- `fix`: Patches a bug in the codebase
|
||||
- `refactor`: Code change that neither fixes a bug nor adds a feature; also applies to removing features
|
||||
- `ci`: Changes to CI configuration
|
||||
- `docs`: Documentation-only changes
|
||||
- `dx`: Improvements to the developer experience
|
||||
|
||||
**Recommended Base Scopes:**
|
||||
|
||||
- `platform`: Changes affecting both frontend and backend
|
||||
- `frontend`
|
||||
- `backend`
|
||||
- `infra`
|
||||
- `blocks`: Modifications/additions of individual blocks
|
||||
|
||||
**Subscope Examples:**
|
||||
|
||||
- `backend/executor`
|
||||
- `backend/db`
|
||||
- `frontend/builder` (includes changes to the block UI component)
|
||||
- `infra/prod`
|
||||
|
||||
Use these scopes and subscopes for clarity and consistency in commit messages.
|
||||
|
||||
@@ -178,7 +178,6 @@ SMTP_USERNAME=
|
||||
SMTP_PASSWORD=
|
||||
|
||||
# Business & Marketing Tools
|
||||
AGENTMAIL_API_KEY=
|
||||
APOLLO_API_KEY=
|
||||
ENRICHLAYER_API_KEY=
|
||||
AYRSHARE_API_KEY=
|
||||
|
||||
@@ -1,227 +0,0 @@
|
||||
# Backend
|
||||
|
||||
This file provides guidance to coding agents when working with the backend.
|
||||
|
||||
## Essential Commands
|
||||
|
||||
To run something with Python package dependencies you MUST use `poetry run ...`.
|
||||
|
||||
```bash
|
||||
# Install dependencies
|
||||
poetry install
|
||||
|
||||
# Run database migrations
|
||||
poetry run prisma migrate dev
|
||||
|
||||
# Start all services (database, redis, rabbitmq, clamav)
|
||||
docker compose up -d
|
||||
|
||||
# Run the backend as a whole
|
||||
poetry run app
|
||||
|
||||
# Run tests
|
||||
poetry run test
|
||||
|
||||
# Run specific test
|
||||
poetry run pytest path/to/test_file.py::test_function_name
|
||||
|
||||
# Run block tests (tests that validate all blocks work correctly)
|
||||
poetry run pytest backend/blocks/test/test_block.py -xvs
|
||||
|
||||
# Run tests for a specific block (e.g., GetCurrentTimeBlock)
|
||||
poetry run pytest 'backend/blocks/test/test_block.py::test_available_blocks[GetCurrentTimeBlock]' -xvs
|
||||
|
||||
# Lint and format
|
||||
# prefer format if you want to just "fix" it and only get the errors that can't be autofixed
|
||||
poetry run format # Black + isort
|
||||
poetry run lint # ruff
|
||||
```
|
||||
|
||||
More details can be found in @TESTING.md
|
||||
|
||||
### Creating/Updating Snapshots
|
||||
|
||||
When you first write a test or when the expected output changes:
|
||||
|
||||
```bash
|
||||
poetry run pytest path/to/test.py --snapshot-update
|
||||
```
|
||||
|
||||
⚠️ **Important**: Always review snapshot changes before committing! Use `git diff` to verify the changes are expected.
|
||||
|
||||
## Architecture
|
||||
|
||||
- **API Layer**: FastAPI with REST and WebSocket endpoints
|
||||
- **Database**: PostgreSQL with Prisma ORM, includes pgvector for embeddings
|
||||
- **Queue System**: RabbitMQ for async task processing
|
||||
- **Execution Engine**: Separate executor service processes agent workflows
|
||||
- **Authentication**: JWT-based with Supabase integration
|
||||
- **Security**: Cache protection middleware prevents sensitive data caching in browsers/proxies
|
||||
|
||||
## Code Style
|
||||
|
||||
- **Top-level imports only** — no local/inner imports (lazy imports only for heavy optional deps like `openpyxl`)
|
||||
- **Absolute imports** — use `from backend.module import ...` for cross-package imports. Single-dot relative (`from .sibling import ...`) is acceptable for sibling modules within the same package (e.g., blocks). Avoid double-dot relative imports (`from ..parent import ...`) — use the absolute path instead
|
||||
- **No duck typing** — no `hasattr`/`getattr`/`isinstance` for type dispatch; use typed interfaces/unions/protocols
|
||||
- **Pydantic models** over dataclass/namedtuple/dict for structured data
|
||||
- **No linter suppressors** — no `# type: ignore`, `# noqa`, `# pyright: ignore`; fix the type/code
|
||||
- **List comprehensions** over manual loop-and-append
|
||||
- **Early return** — guard clauses first, avoid deep nesting
|
||||
- **f-strings vs printf syntax in log statements** — Use `%s` for deferred interpolation in `debug` statements, f-strings elsewhere for readability: `logger.debug("Processing %s items", count)`, `logger.info(f"Processing {count} items")`
|
||||
- **Sanitize error paths** — `os.path.basename()` in error messages to avoid leaking directory structure
|
||||
- **TOCTOU awareness** — avoid check-then-act patterns for file access and credit charging
|
||||
- **`Security()` vs `Depends()`** — use `Security()` for auth deps to get proper OpenAPI security spec
|
||||
- **Redis pipelines** — `transaction=True` for atomicity on multi-step operations
|
||||
- **`max(0, value)` guards** — for computed values that should never be negative
|
||||
- **SSE protocol** — `data:` lines for frontend-parsed events (must match Zod schema), `: comment` lines for heartbeats/status
|
||||
- **File length** — keep files under ~300 lines; if a file grows beyond this, split by responsibility (e.g. extract helpers, models, or a sub-module into a new file). Never keep appending to a long file.
|
||||
- **Function length** — keep functions under ~40 lines; extract named helpers when a function grows longer. Long functions are a sign of mixed concerns, not complexity.
|
||||
- **Top-down ordering** — define the main/public function or class first, then the helpers it uses below. A reader should encounter high-level logic before implementation details.
|
||||
|
||||
## Testing Approach
|
||||
|
||||
- Uses pytest with snapshot testing for API responses
|
||||
- Test files are colocated with source files (`*_test.py`)
|
||||
- Mock at boundaries — mock where the symbol is **used**, not where it's **defined**
|
||||
- After refactoring, update mock targets to match new module paths
|
||||
- Use `AsyncMock` for async functions (`from unittest.mock import AsyncMock`)
|
||||
|
||||
### Test-Driven Development (TDD)
|
||||
|
||||
When fixing a bug or adding a feature, write the test **before** the implementation:
|
||||
|
||||
```python
|
||||
# 1. Write a failing test marked xfail
|
||||
@pytest.mark.xfail(reason="Bug #1234: widget crashes on empty input")
|
||||
def test_widget_handles_empty_input():
|
||||
result = widget.process("")
|
||||
assert result == Widget.EMPTY_RESULT
|
||||
|
||||
# 2. Run it — confirm it fails (XFAIL)
|
||||
# poetry run pytest path/to/test.py::test_widget_handles_empty_input -xvs
|
||||
|
||||
# 3. Implement the fix
|
||||
|
||||
# 4. Remove xfail, run again — confirm it passes
|
||||
def test_widget_handles_empty_input():
|
||||
result = widget.process("")
|
||||
assert result == Widget.EMPTY_RESULT
|
||||
```
|
||||
|
||||
This catches regressions and proves the fix actually works. **Every bug fix should include a test that would have caught it.**
|
||||
|
||||
## Database Schema
|
||||
|
||||
Key models (defined in `schema.prisma`):
|
||||
|
||||
- `User`: Authentication and profile data
|
||||
- `AgentGraph`: Workflow definitions with version control
|
||||
- `AgentGraphExecution`: Execution history and results
|
||||
- `AgentNode`: Individual nodes in a workflow
|
||||
- `StoreListing`: Marketplace listings for sharing agents
|
||||
|
||||
## Environment Configuration
|
||||
|
||||
- **Backend**: `.env.default` (defaults) → `.env` (user overrides)
|
||||
|
||||
## Common Development Tasks
|
||||
|
||||
### Adding a new block
|
||||
|
||||
Follow the comprehensive [Block SDK Guide](@../../docs/platform/block-sdk-guide.md) which covers:
|
||||
|
||||
- Provider configuration with `ProviderBuilder`
|
||||
- Block schema definition
|
||||
- Authentication (API keys, OAuth, webhooks)
|
||||
- Testing and validation
|
||||
- File organization
|
||||
|
||||
Quick steps:
|
||||
|
||||
1. Create new file in `backend/blocks/`
|
||||
2. Configure provider using `ProviderBuilder` in `_config.py`
|
||||
3. Inherit from `Block` base class
|
||||
4. Define input/output schemas using `BlockSchema`
|
||||
5. Implement async `run` method
|
||||
6. Generate unique block ID using `uuid.uuid4()`
|
||||
7. Test with `poetry run pytest backend/blocks/test/test_block.py`
|
||||
|
||||
Note: when making many new blocks analyze the interfaces for each of these blocks and picture if they would go well together in a graph-based editor or would they struggle to connect productively?
|
||||
ex: do the inputs and outputs tie well together?
|
||||
|
||||
If you get any pushback or hit complex block conditions check the new_blocks guide in the docs.
|
||||
|
||||
#### Handling files in blocks with `store_media_file()`
|
||||
|
||||
When blocks need to work with files (images, videos, documents), use `store_media_file()` from `backend.util.file`. The `return_format` parameter determines what you get back:
|
||||
|
||||
| Format | Use When | Returns |
|
||||
|--------|----------|---------|
|
||||
| `"for_local_processing"` | Processing with local tools (ffmpeg, MoviePy, PIL) | Local file path (e.g., `"image.png"`) |
|
||||
| `"for_external_api"` | Sending content to external APIs (Replicate, OpenAI) | Data URI (e.g., `"data:image/png;base64,..."`) |
|
||||
| `"for_block_output"` | Returning output from your block | Smart: `workspace://` in CoPilot, data URI in graphs |
|
||||
|
||||
**Examples:**
|
||||
|
||||
```python
|
||||
# INPUT: Need to process file locally with ffmpeg
|
||||
local_path = await store_media_file(
|
||||
file=input_data.video,
|
||||
execution_context=execution_context,
|
||||
return_format="for_local_processing",
|
||||
)
|
||||
# local_path = "video.mp4" - use with Path/ffmpeg/etc
|
||||
|
||||
# INPUT: Need to send to external API like Replicate
|
||||
image_b64 = await store_media_file(
|
||||
file=input_data.image,
|
||||
execution_context=execution_context,
|
||||
return_format="for_external_api",
|
||||
)
|
||||
# image_b64 = "data:image/png;base64,iVBORw0..." - send to API
|
||||
|
||||
# OUTPUT: Returning result from block
|
||||
result_url = await store_media_file(
|
||||
file=generated_image_url,
|
||||
execution_context=execution_context,
|
||||
return_format="for_block_output",
|
||||
)
|
||||
yield "image_url", result_url
|
||||
# In CoPilot: result_url = "workspace://abc123"
|
||||
# In graphs: result_url = "data:image/png;base64,..."
|
||||
```
|
||||
|
||||
**Key points:**
|
||||
|
||||
- `for_block_output` is the ONLY format that auto-adapts to execution context
|
||||
- Always use `for_block_output` for block outputs unless you have a specific reason not to
|
||||
- Never hardcode workspace checks - let `for_block_output` handle it
|
||||
|
||||
### Modifying the API
|
||||
|
||||
1. Update route in `backend/api/features/`
|
||||
2. Add/update Pydantic models in same directory
|
||||
3. Write tests alongside the route file
|
||||
4. Run `poetry run test` to verify
|
||||
|
||||
## Workspace & Media Files
|
||||
|
||||
**Read [Workspace & Media Architecture](../../docs/platform/workspace-media-architecture.md) when:**
|
||||
- Working on CoPilot file upload/download features
|
||||
- Building blocks that handle `MediaFileType` inputs/outputs
|
||||
- Modifying `WorkspaceManager` or `store_media_file()`
|
||||
- Debugging file persistence or virus scanning issues
|
||||
|
||||
Covers: `WorkspaceManager` (persistent storage with session scoping), `store_media_file()` (media normalization pipeline), and responsibility boundaries for virus scanning and persistence.
|
||||
|
||||
## Security Implementation
|
||||
|
||||
### Cache Protection Middleware
|
||||
|
||||
- Located in `backend/api/middleware/security.py`
|
||||
- Default behavior: Disables caching for ALL endpoints with `Cache-Control: no-store, no-cache, must-revalidate, private`
|
||||
- Uses an allow list approach - only explicitly permitted paths can be cached
|
||||
- Cacheable paths include: static assets (`static/*`, `_next/static/*`), health checks, public store pages, documentation
|
||||
- Prevents sensitive data (auth tokens, API keys, user data) from being cached by browsers/proxies
|
||||
- To allow caching for a new endpoint, add it to `CACHEABLE_PATHS` in the middleware
|
||||
- Applied to both main API server and external API applications
|
||||
@@ -1 +1,227 @@
|
||||
@AGENTS.md
|
||||
# CLAUDE.md - Backend
|
||||
|
||||
This file provides guidance to Claude Code when working with the backend.
|
||||
|
||||
## Essential Commands
|
||||
|
||||
To run something with Python package dependencies you MUST use `poetry run ...`.
|
||||
|
||||
```bash
|
||||
# Install dependencies
|
||||
poetry install
|
||||
|
||||
# Run database migrations
|
||||
poetry run prisma migrate dev
|
||||
|
||||
# Start all services (database, redis, rabbitmq, clamav)
|
||||
docker compose up -d
|
||||
|
||||
# Run the backend as a whole
|
||||
poetry run app
|
||||
|
||||
# Run tests
|
||||
poetry run test
|
||||
|
||||
# Run specific test
|
||||
poetry run pytest path/to/test_file.py::test_function_name
|
||||
|
||||
# Run block tests (tests that validate all blocks work correctly)
|
||||
poetry run pytest backend/blocks/test/test_block.py -xvs
|
||||
|
||||
# Run tests for a specific block (e.g., GetCurrentTimeBlock)
|
||||
poetry run pytest 'backend/blocks/test/test_block.py::test_available_blocks[GetCurrentTimeBlock]' -xvs
|
||||
|
||||
# Lint and format
|
||||
# prefer format if you want to just "fix" it and only get the errors that can't be autofixed
|
||||
poetry run format # Black + isort
|
||||
poetry run lint # ruff
|
||||
```
|
||||
|
||||
More details can be found in @TESTING.md
|
||||
|
||||
### Creating/Updating Snapshots
|
||||
|
||||
When you first write a test or when the expected output changes:
|
||||
|
||||
```bash
|
||||
poetry run pytest path/to/test.py --snapshot-update
|
||||
```
|
||||
|
||||
⚠️ **Important**: Always review snapshot changes before committing! Use `git diff` to verify the changes are expected.
|
||||
|
||||
## Architecture
|
||||
|
||||
- **API Layer**: FastAPI with REST and WebSocket endpoints
|
||||
- **Database**: PostgreSQL with Prisma ORM, includes pgvector for embeddings
|
||||
- **Queue System**: RabbitMQ for async task processing
|
||||
- **Execution Engine**: Separate executor service processes agent workflows
|
||||
- **Authentication**: JWT-based with Supabase integration
|
||||
- **Security**: Cache protection middleware prevents sensitive data caching in browsers/proxies
|
||||
|
||||
## Code Style
|
||||
|
||||
- **Top-level imports only** — no local/inner imports (lazy imports only for heavy optional deps like `openpyxl`)
|
||||
- **Absolute imports** — use `from backend.module import ...` for cross-package imports. Single-dot relative (`from .sibling import ...`) is acceptable for sibling modules within the same package (e.g., blocks). Avoid double-dot relative imports (`from ..parent import ...`) — use the absolute path instead
|
||||
- **No duck typing** — no `hasattr`/`getattr`/`isinstance` for type dispatch; use typed interfaces/unions/protocols
|
||||
- **Pydantic models** over dataclass/namedtuple/dict for structured data
|
||||
- **No linter suppressors** — no `# type: ignore`, `# noqa`, `# pyright: ignore`; fix the type/code
|
||||
- **List comprehensions** over manual loop-and-append
|
||||
- **Early return** — guard clauses first, avoid deep nesting
|
||||
- **f-strings vs printf syntax in log statements** — Use `%s` for deferred interpolation in `debug` statements, f-strings elsewhere for readability: `logger.debug("Processing %s items", count)`, `logger.info(f"Processing {count} items")`
|
||||
- **Sanitize error paths** — `os.path.basename()` in error messages to avoid leaking directory structure
|
||||
- **TOCTOU awareness** — avoid check-then-act patterns for file access and credit charging
|
||||
- **`Security()` vs `Depends()`** — use `Security()` for auth deps to get proper OpenAPI security spec
|
||||
- **Redis pipelines** — `transaction=True` for atomicity on multi-step operations
|
||||
- **`max(0, value)` guards** — for computed values that should never be negative
|
||||
- **SSE protocol** — `data:` lines for frontend-parsed events (must match Zod schema), `: comment` lines for heartbeats/status
|
||||
- **File length** — keep files under ~300 lines; if a file grows beyond this, split by responsibility (e.g. extract helpers, models, or a sub-module into a new file). Never keep appending to a long file.
|
||||
- **Function length** — keep functions under ~40 lines; extract named helpers when a function grows longer. Long functions are a sign of mixed concerns, not complexity.
|
||||
- **Top-down ordering** — define the main/public function or class first, then the helpers it uses below. A reader should encounter high-level logic before implementation details.
|
||||
|
||||
## Testing Approach
|
||||
|
||||
- Uses pytest with snapshot testing for API responses
|
||||
- Test files are colocated with source files (`*_test.py`)
|
||||
- Mock at boundaries — mock where the symbol is **used**, not where it's **defined**
|
||||
- After refactoring, update mock targets to match new module paths
|
||||
- Use `AsyncMock` for async functions (`from unittest.mock import AsyncMock`)
|
||||
|
||||
### Test-Driven Development (TDD)
|
||||
|
||||
When fixing a bug or adding a feature, write the test **before** the implementation:
|
||||
|
||||
```python
|
||||
# 1. Write a failing test marked xfail
|
||||
@pytest.mark.xfail(reason="Bug #1234: widget crashes on empty input")
|
||||
def test_widget_handles_empty_input():
|
||||
result = widget.process("")
|
||||
assert result == Widget.EMPTY_RESULT
|
||||
|
||||
# 2. Run it — confirm it fails (XFAIL)
|
||||
# poetry run pytest path/to/test.py::test_widget_handles_empty_input -xvs
|
||||
|
||||
# 3. Implement the fix
|
||||
|
||||
# 4. Remove xfail, run again — confirm it passes
|
||||
def test_widget_handles_empty_input():
|
||||
result = widget.process("")
|
||||
assert result == Widget.EMPTY_RESULT
|
||||
```
|
||||
|
||||
This catches regressions and proves the fix actually works. **Every bug fix should include a test that would have caught it.**
|
||||
|
||||
## Database Schema
|
||||
|
||||
Key models (defined in `schema.prisma`):
|
||||
|
||||
- `User`: Authentication and profile data
|
||||
- `AgentGraph`: Workflow definitions with version control
|
||||
- `AgentGraphExecution`: Execution history and results
|
||||
- `AgentNode`: Individual nodes in a workflow
|
||||
- `StoreListing`: Marketplace listings for sharing agents
|
||||
|
||||
## Environment Configuration
|
||||
|
||||
- **Backend**: `.env.default` (defaults) → `.env` (user overrides)
|
||||
|
||||
## Common Development Tasks
|
||||
|
||||
### Adding a new block
|
||||
|
||||
Follow the comprehensive [Block SDK Guide](@../../docs/content/platform/block-sdk-guide.md) which covers:
|
||||
|
||||
- Provider configuration with `ProviderBuilder`
|
||||
- Block schema definition
|
||||
- Authentication (API keys, OAuth, webhooks)
|
||||
- Testing and validation
|
||||
- File organization
|
||||
|
||||
Quick steps:
|
||||
|
||||
1. Create new file in `backend/blocks/`
|
||||
2. Configure provider using `ProviderBuilder` in `_config.py`
|
||||
3. Inherit from `Block` base class
|
||||
4. Define input/output schemas using `BlockSchema`
|
||||
5. Implement async `run` method
|
||||
6. Generate unique block ID using `uuid.uuid4()`
|
||||
7. Test with `poetry run pytest backend/blocks/test/test_block.py`
|
||||
|
||||
Note: when making many new blocks analyze the interfaces for each of these blocks and picture if they would go well together in a graph-based editor or would they struggle to connect productively?
|
||||
ex: do the inputs and outputs tie well together?
|
||||
|
||||
If you get any pushback or hit complex block conditions check the new_blocks guide in the docs.
|
||||
|
||||
#### Handling files in blocks with `store_media_file()`
|
||||
|
||||
When blocks need to work with files (images, videos, documents), use `store_media_file()` from `backend.util.file`. The `return_format` parameter determines what you get back:
|
||||
|
||||
| Format | Use When | Returns |
|
||||
|--------|----------|---------|
|
||||
| `"for_local_processing"` | Processing with local tools (ffmpeg, MoviePy, PIL) | Local file path (e.g., `"image.png"`) |
|
||||
| `"for_external_api"` | Sending content to external APIs (Replicate, OpenAI) | Data URI (e.g., `"data:image/png;base64,..."`) |
|
||||
| `"for_block_output"` | Returning output from your block | Smart: `workspace://` in CoPilot, data URI in graphs |
|
||||
|
||||
**Examples:**
|
||||
|
||||
```python
|
||||
# INPUT: Need to process file locally with ffmpeg
|
||||
local_path = await store_media_file(
|
||||
file=input_data.video,
|
||||
execution_context=execution_context,
|
||||
return_format="for_local_processing",
|
||||
)
|
||||
# local_path = "video.mp4" - use with Path/ffmpeg/etc
|
||||
|
||||
# INPUT: Need to send to external API like Replicate
|
||||
image_b64 = await store_media_file(
|
||||
file=input_data.image,
|
||||
execution_context=execution_context,
|
||||
return_format="for_external_api",
|
||||
)
|
||||
# image_b64 = "data:image/png;base64,iVBORw0..." - send to API
|
||||
|
||||
# OUTPUT: Returning result from block
|
||||
result_url = await store_media_file(
|
||||
file=generated_image_url,
|
||||
execution_context=execution_context,
|
||||
return_format="for_block_output",
|
||||
)
|
||||
yield "image_url", result_url
|
||||
# In CoPilot: result_url = "workspace://abc123"
|
||||
# In graphs: result_url = "data:image/png;base64,..."
|
||||
```
|
||||
|
||||
**Key points:**
|
||||
|
||||
- `for_block_output` is the ONLY format that auto-adapts to execution context
|
||||
- Always use `for_block_output` for block outputs unless you have a specific reason not to
|
||||
- Never hardcode workspace checks - let `for_block_output` handle it
|
||||
|
||||
### Modifying the API
|
||||
|
||||
1. Update route in `backend/api/features/`
|
||||
2. Add/update Pydantic models in same directory
|
||||
3. Write tests alongside the route file
|
||||
4. Run `poetry run test` to verify
|
||||
|
||||
## Workspace & Media Files
|
||||
|
||||
**Read [Workspace & Media Architecture](../../docs/platform/workspace-media-architecture.md) when:**
|
||||
- Working on CoPilot file upload/download features
|
||||
- Building blocks that handle `MediaFileType` inputs/outputs
|
||||
- Modifying `WorkspaceManager` or `store_media_file()`
|
||||
- Debugging file persistence or virus scanning issues
|
||||
|
||||
Covers: `WorkspaceManager` (persistent storage with session scoping), `store_media_file()` (media normalization pipeline), and responsibility boundaries for virus scanning and persistence.
|
||||
|
||||
## Security Implementation
|
||||
|
||||
### Cache Protection Middleware
|
||||
|
||||
- Located in `backend/api/middleware/security.py`
|
||||
- Default behavior: Disables caching for ALL endpoints with `Cache-Control: no-store, no-cache, must-revalidate, private`
|
||||
- Uses an allow list approach - only explicitly permitted paths can be cached
|
||||
- Cacheable paths include: static assets (`static/*`, `_next/static/*`), health checks, public store pages, documentation
|
||||
- Prevents sensitive data (auth tokens, API keys, user data) from being cached by browsers/proxies
|
||||
- To allow caching for a new endpoint, add it to `CACHEABLE_PATHS` in the middleware
|
||||
- Applied to both main API server and external API applications
|
||||
|
||||
@@ -31,10 +31,7 @@ from backend.data.model import (
|
||||
UserPasswordCredentials,
|
||||
is_sdk_default,
|
||||
)
|
||||
from backend.integrations.credentials_store import (
|
||||
is_system_credential,
|
||||
provider_matches,
|
||||
)
|
||||
from backend.integrations.credentials_store import provider_matches
|
||||
from backend.integrations.creds_manager import IntegrationCredentialsManager
|
||||
from backend.integrations.oauth import CREDENTIALS_BY_PROVIDER, HANDLERS_BY_NAME
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -621,11 +618,6 @@ async def delete_credential(
|
||||
raise HTTPException(
|
||||
status_code=status.HTTP_404_NOT_FOUND, detail="Credentials not found"
|
||||
)
|
||||
if is_system_credential(cred_id):
|
||||
raise HTTPException(
|
||||
status_code=status.HTTP_403_FORBIDDEN,
|
||||
detail="System-managed credentials cannot be deleted",
|
||||
)
|
||||
creds = await creds_manager.store.get_creds_by_id(auth.user_id, cred_id)
|
||||
if not creds:
|
||||
raise HTTPException(
|
||||
|
||||
@@ -72,7 +72,7 @@ class RunAgentRequest(BaseModel):
|
||||
|
||||
def _create_ephemeral_session(user_id: str) -> ChatSession:
|
||||
"""Create an ephemeral session for stateless API requests."""
|
||||
return ChatSession.new(user_id, dry_run=False)
|
||||
return ChatSession.new(user_id)
|
||||
|
||||
|
||||
@tools_router.post(
|
||||
|
||||
@@ -1,85 +0,0 @@
|
||||
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,
|
||||
),
|
||||
)
|
||||
@@ -1,135 +0,0 @@
|
||||
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)
|
||||
@@ -1,253 +0,0 @@
|
||||
"""Admin endpoints for checking and resetting user CoPilot rate limit usage."""
|
||||
|
||||
import logging
|
||||
from typing import Optional
|
||||
|
||||
from autogpt_libs.auth import get_user_id, requires_admin_user
|
||||
from fastapi import APIRouter, Body, HTTPException, Security
|
||||
from pydantic import BaseModel
|
||||
|
||||
from backend.copilot.config import ChatConfig
|
||||
from backend.copilot.rate_limit import (
|
||||
SubscriptionTier,
|
||||
get_global_rate_limits,
|
||||
get_usage_status,
|
||||
get_user_tier,
|
||||
reset_user_usage,
|
||||
set_user_tier,
|
||||
)
|
||||
from backend.data.user import get_user_by_email, get_user_email_by_id, search_users
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
config = ChatConfig()
|
||||
|
||||
router = APIRouter(
|
||||
prefix="/admin",
|
||||
tags=["copilot", "admin"],
|
||||
dependencies=[Security(requires_admin_user)],
|
||||
)
|
||||
|
||||
|
||||
class UserRateLimitResponse(BaseModel):
|
||||
user_id: str
|
||||
user_email: Optional[str] = None
|
||||
daily_token_limit: int
|
||||
weekly_token_limit: int
|
||||
daily_tokens_used: int
|
||||
weekly_tokens_used: int
|
||||
tier: SubscriptionTier
|
||||
|
||||
|
||||
class UserTierResponse(BaseModel):
|
||||
user_id: str
|
||||
tier: SubscriptionTier
|
||||
|
||||
|
||||
class SetUserTierRequest(BaseModel):
|
||||
user_id: str
|
||||
tier: SubscriptionTier
|
||||
|
||||
|
||||
async def _resolve_user_id(
|
||||
user_id: Optional[str], email: Optional[str]
|
||||
) -> tuple[str, Optional[str]]:
|
||||
"""Resolve a user_id and email from the provided parameters.
|
||||
|
||||
Returns (user_id, email). Accepts either user_id or email; at least one
|
||||
must be provided. When both are provided, ``email`` takes precedence.
|
||||
"""
|
||||
if email:
|
||||
user = await get_user_by_email(email)
|
||||
if not user:
|
||||
raise HTTPException(
|
||||
status_code=404, detail="No user found with the provided email."
|
||||
)
|
||||
return user.id, email
|
||||
|
||||
if not user_id:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="Either user_id or email query parameter is required.",
|
||||
)
|
||||
|
||||
# We have a user_id; try to look up their email for display purposes.
|
||||
# This is non-critical -- a failure should not block the response.
|
||||
try:
|
||||
resolved_email = await get_user_email_by_id(user_id)
|
||||
except Exception:
|
||||
logger.warning("Failed to resolve email for user %s", user_id, exc_info=True)
|
||||
resolved_email = None
|
||||
return user_id, resolved_email
|
||||
|
||||
|
||||
@router.get(
|
||||
"/rate_limit",
|
||||
response_model=UserRateLimitResponse,
|
||||
summary="Get User Rate Limit",
|
||||
)
|
||||
async def get_user_rate_limit(
|
||||
user_id: Optional[str] = None,
|
||||
email: Optional[str] = None,
|
||||
admin_user_id: str = Security(get_user_id),
|
||||
) -> UserRateLimitResponse:
|
||||
"""Get a user's current usage and effective rate limits. Admin-only.
|
||||
|
||||
Accepts either ``user_id`` or ``email`` as a query parameter.
|
||||
When ``email`` is provided the user is looked up by email first.
|
||||
"""
|
||||
resolved_id, resolved_email = await _resolve_user_id(user_id, email)
|
||||
|
||||
logger.info("Admin %s checking rate limit for user %s", admin_user_id, resolved_id)
|
||||
|
||||
daily_limit, weekly_limit, tier = await get_global_rate_limits(
|
||||
resolved_id, config.daily_token_limit, config.weekly_token_limit
|
||||
)
|
||||
usage = await get_usage_status(resolved_id, daily_limit, weekly_limit, tier=tier)
|
||||
|
||||
return UserRateLimitResponse(
|
||||
user_id=resolved_id,
|
||||
user_email=resolved_email,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
daily_tokens_used=usage.daily.used,
|
||||
weekly_tokens_used=usage.weekly.used,
|
||||
tier=tier,
|
||||
)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/rate_limit/reset",
|
||||
response_model=UserRateLimitResponse,
|
||||
summary="Reset User Rate Limit Usage",
|
||||
)
|
||||
async def reset_user_rate_limit(
|
||||
user_id: str = Body(embed=True),
|
||||
reset_weekly: bool = Body(False, embed=True),
|
||||
admin_user_id: str = Security(get_user_id),
|
||||
) -> UserRateLimitResponse:
|
||||
"""Reset a user's daily usage counter (and optionally weekly). Admin-only."""
|
||||
logger.info(
|
||||
"Admin %s resetting rate limit for user %s (reset_weekly=%s)",
|
||||
admin_user_id,
|
||||
user_id,
|
||||
reset_weekly,
|
||||
)
|
||||
|
||||
try:
|
||||
await reset_user_usage(user_id, reset_weekly=reset_weekly)
|
||||
except Exception as e:
|
||||
logger.exception("Failed to reset user usage")
|
||||
raise HTTPException(status_code=500, detail="Failed to reset usage") from e
|
||||
|
||||
daily_limit, weekly_limit, tier = await get_global_rate_limits(
|
||||
user_id, config.daily_token_limit, config.weekly_token_limit
|
||||
)
|
||||
usage = await get_usage_status(user_id, daily_limit, weekly_limit, tier=tier)
|
||||
|
||||
try:
|
||||
resolved_email = await get_user_email_by_id(user_id)
|
||||
except Exception:
|
||||
logger.warning("Failed to resolve email for user %s", user_id, exc_info=True)
|
||||
resolved_email = None
|
||||
|
||||
return UserRateLimitResponse(
|
||||
user_id=user_id,
|
||||
user_email=resolved_email,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
daily_tokens_used=usage.daily.used,
|
||||
weekly_tokens_used=usage.weekly.used,
|
||||
tier=tier,
|
||||
)
|
||||
|
||||
|
||||
@router.get(
|
||||
"/rate_limit/tier",
|
||||
response_model=UserTierResponse,
|
||||
summary="Get User Rate Limit Tier",
|
||||
)
|
||||
async def get_user_rate_limit_tier(
|
||||
user_id: str,
|
||||
admin_user_id: str = Security(get_user_id),
|
||||
) -> UserTierResponse:
|
||||
"""Get a user's current rate-limit tier. Admin-only.
|
||||
|
||||
Returns 404 if the user does not exist in the database.
|
||||
"""
|
||||
logger.info("Admin %s checking tier for user %s", admin_user_id, user_id)
|
||||
|
||||
resolved_email = await get_user_email_by_id(user_id)
|
||||
if resolved_email is None:
|
||||
raise HTTPException(status_code=404, detail=f"User {user_id} not found")
|
||||
|
||||
tier = await get_user_tier(user_id)
|
||||
return UserTierResponse(user_id=user_id, tier=tier)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/rate_limit/tier",
|
||||
response_model=UserTierResponse,
|
||||
summary="Set User Rate Limit Tier",
|
||||
)
|
||||
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."""
|
||||
old_tier = await get_user_tier(request.user_id)
|
||||
|
||||
# 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
|
||||
)
|
||||
resolved_email = None
|
||||
logger.info(
|
||||
"Admin %s changing tier for user %s (%s): %s -> %s",
|
||||
admin_user_id,
|
||||
request.user_id,
|
||||
resolved_email or "unknown",
|
||||
old_tier.value,
|
||||
request.tier.value,
|
||||
)
|
||||
try:
|
||||
await set_user_tier(request.user_id, request.tier)
|
||||
except Exception as e:
|
||||
logger.exception("Failed to set user tier")
|
||||
raise HTTPException(status_code=500, detail="Failed to set tier") from e
|
||||
|
||||
return UserTierResponse(user_id=request.user_id, tier=request.tier)
|
||||
|
||||
|
||||
class UserSearchResult(BaseModel):
|
||||
user_id: str
|
||||
user_email: Optional[str] = None
|
||||
|
||||
|
||||
@router.get(
|
||||
"/rate_limit/search_users",
|
||||
response_model=list[UserSearchResult],
|
||||
summary="Search Users by Name or Email",
|
||||
)
|
||||
async def admin_search_users(
|
||||
query: str,
|
||||
limit: int = 20,
|
||||
admin_user_id: str = Security(get_user_id),
|
||||
) -> list[UserSearchResult]:
|
||||
"""Search users by partial email or name. Admin-only.
|
||||
|
||||
Queries the User table directly — returns results even for users
|
||||
without credit transaction history.
|
||||
"""
|
||||
if len(query.strip()) < 3:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="Search query must be at least 3 characters.",
|
||||
)
|
||||
logger.info("Admin %s searching users with query=%r", admin_user_id, query)
|
||||
results = await search_users(query, limit=max(1, min(limit, 50)))
|
||||
return [UserSearchResult(user_id=uid, user_email=email) for uid, email in results]
|
||||
@@ -1,557 +0,0 @@
|
||||
import json
|
||||
from types import SimpleNamespace
|
||||
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 pytest_snapshot.plugin import Snapshot
|
||||
|
||||
from backend.copilot.rate_limit import CoPilotUsageStatus, SubscriptionTier, UsageWindow
|
||||
|
||||
from .rate_limit_admin_routes import router as rate_limit_admin_router
|
||||
|
||||
app = fastapi.FastAPI()
|
||||
app.include_router(rate_limit_admin_router)
|
||||
|
||||
client = fastapi.testclient.TestClient(app)
|
||||
|
||||
_MOCK_MODULE = "backend.api.features.admin.rate_limit_admin_routes"
|
||||
|
||||
_TARGET_EMAIL = "target@example.com"
|
||||
|
||||
|
||||
@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 _mock_usage_status(
|
||||
daily_used: int = 500_000, weekly_used: int = 3_000_000
|
||||
) -> CoPilotUsageStatus:
|
||||
from datetime import UTC, datetime, timedelta
|
||||
|
||||
now = datetime.now(UTC)
|
||||
return CoPilotUsageStatus(
|
||||
daily=UsageWindow(
|
||||
used=daily_used, limit=2_500_000, resets_at=now + timedelta(hours=6)
|
||||
),
|
||||
weekly=UsageWindow(
|
||||
used=weekly_used, limit=12_500_000, resets_at=now + timedelta(days=3)
|
||||
),
|
||||
)
|
||||
|
||||
|
||||
def _patch_rate_limit_deps(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
daily_used: int = 500_000,
|
||||
weekly_used: int = 3_000_000,
|
||||
):
|
||||
"""Patch the common rate-limit + user-lookup dependencies."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_global_rate_limits",
|
||||
new_callable=AsyncMock,
|
||||
return_value=(2_500_000, 12_500_000, SubscriptionTier.FREE),
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_usage_status",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_mock_usage_status(daily_used=daily_used, weekly_used=weekly_used),
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_TARGET_EMAIL,
|
||||
)
|
||||
|
||||
|
||||
def test_get_rate_limit(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
configured_snapshot: Snapshot,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test getting rate limit and usage for a user."""
|
||||
_patch_rate_limit_deps(mocker, target_user_id)
|
||||
|
||||
response = client.get("/admin/rate_limit", params={"user_id": target_user_id})
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["user_email"] == _TARGET_EMAIL
|
||||
assert data["daily_token_limit"] == 2_500_000
|
||||
assert data["weekly_token_limit"] == 12_500_000
|
||||
assert data["daily_tokens_used"] == 500_000
|
||||
assert data["weekly_tokens_used"] == 3_000_000
|
||||
assert data["tier"] == "FREE"
|
||||
|
||||
configured_snapshot.assert_match(
|
||||
json.dumps(data, indent=2, sort_keys=True) + "\n",
|
||||
"get_rate_limit",
|
||||
)
|
||||
|
||||
|
||||
def test_get_rate_limit_by_email(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test looking up rate limits via email instead of user_id."""
|
||||
_patch_rate_limit_deps(mocker, target_user_id)
|
||||
|
||||
mock_user = SimpleNamespace(id=target_user_id, email=_TARGET_EMAIL)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_by_email",
|
||||
new_callable=AsyncMock,
|
||||
return_value=mock_user,
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit", params={"email": _TARGET_EMAIL})
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["user_email"] == _TARGET_EMAIL
|
||||
assert data["daily_token_limit"] == 2_500_000
|
||||
|
||||
|
||||
def test_get_rate_limit_by_email_not_found(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
) -> None:
|
||||
"""Test that looking up a non-existent email returns 404."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_by_email",
|
||||
new_callable=AsyncMock,
|
||||
return_value=None,
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit", params={"email": "nobody@example.com"})
|
||||
|
||||
assert response.status_code == 404
|
||||
|
||||
|
||||
def test_get_rate_limit_no_params() -> None:
|
||||
"""Test that omitting both user_id and email returns 400."""
|
||||
response = client.get("/admin/rate_limit")
|
||||
assert response.status_code == 400
|
||||
|
||||
|
||||
def test_reset_user_usage_daily_only(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
configured_snapshot: Snapshot,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test resetting only daily usage (default behaviour)."""
|
||||
mock_reset = mocker.patch(
|
||||
f"{_MOCK_MODULE}.reset_user_usage",
|
||||
new_callable=AsyncMock,
|
||||
)
|
||||
_patch_rate_limit_deps(mocker, target_user_id, daily_used=0, weekly_used=3_000_000)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/reset",
|
||||
json={"user_id": target_user_id},
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["daily_tokens_used"] == 0
|
||||
# Weekly is untouched
|
||||
assert data["weekly_tokens_used"] == 3_000_000
|
||||
assert data["tier"] == "FREE"
|
||||
|
||||
mock_reset.assert_awaited_once_with(target_user_id, reset_weekly=False)
|
||||
|
||||
configured_snapshot.assert_match(
|
||||
json.dumps(data, indent=2, sort_keys=True) + "\n",
|
||||
"reset_user_usage_daily_only",
|
||||
)
|
||||
|
||||
|
||||
def test_reset_user_usage_daily_and_weekly(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
configured_snapshot: Snapshot,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test resetting both daily and weekly usage."""
|
||||
mock_reset = mocker.patch(
|
||||
f"{_MOCK_MODULE}.reset_user_usage",
|
||||
new_callable=AsyncMock,
|
||||
)
|
||||
_patch_rate_limit_deps(mocker, target_user_id, daily_used=0, weekly_used=0)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/reset",
|
||||
json={"user_id": target_user_id, "reset_weekly": True},
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["daily_tokens_used"] == 0
|
||||
assert data["weekly_tokens_used"] == 0
|
||||
assert data["tier"] == "FREE"
|
||||
|
||||
mock_reset.assert_awaited_once_with(target_user_id, reset_weekly=True)
|
||||
|
||||
configured_snapshot.assert_match(
|
||||
json.dumps(data, indent=2, sort_keys=True) + "\n",
|
||||
"reset_user_usage_daily_and_weekly",
|
||||
)
|
||||
|
||||
|
||||
def test_reset_user_usage_redis_failure(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that Redis failure on reset returns 500."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.reset_user_usage",
|
||||
new_callable=AsyncMock,
|
||||
side_effect=Exception("Redis connection refused"),
|
||||
)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/reset",
|
||||
json={"user_id": target_user_id},
|
||||
)
|
||||
|
||||
assert response.status_code == 500
|
||||
|
||||
|
||||
def test_get_rate_limit_email_lookup_failure(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that failing to resolve a user email degrades gracefully."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_global_rate_limits",
|
||||
new_callable=AsyncMock,
|
||||
return_value=(2_500_000, 12_500_000, SubscriptionTier.FREE),
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_usage_status",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_mock_usage_status(),
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
side_effect=Exception("DB connection lost"),
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit", params={"user_id": target_user_id})
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["user_email"] is None
|
||||
|
||||
|
||||
def test_admin_endpoints_require_admin_role(mock_jwt_user) -> None:
|
||||
"""Test that rate limit admin endpoints require admin role."""
|
||||
app.dependency_overrides[get_jwt_payload] = mock_jwt_user["get_jwt_payload"]
|
||||
|
||||
response = client.get("/admin/rate_limit", params={"user_id": "test"})
|
||||
assert response.status_code == 403
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/reset",
|
||||
json={"user_id": "test"},
|
||||
)
|
||||
assert response.status_code == 403
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Tier management endpoints
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
def test_get_user_tier(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test getting a user's rate-limit tier."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_TARGET_EMAIL,
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit/tier", params={"user_id": target_user_id})
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["tier"] == "PRO"
|
||||
|
||||
|
||||
def test_get_user_tier_user_not_found(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that getting tier for a non-existent user returns 404."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=None,
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit/tier", params={"user_id": target_user_id})
|
||||
|
||||
assert response.status_code == 404
|
||||
|
||||
|
||||
def test_set_user_tier(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test setting a user's rate-limit tier (upgrade)."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_TARGET_EMAIL,
|
||||
)
|
||||
mocker.patch(
|
||||
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,
|
||||
)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "ENTERPRISE"},
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["tier"] == "ENTERPRISE"
|
||||
mock_set.assert_awaited_once_with(target_user_id, SubscriptionTier.ENTERPRISE)
|
||||
|
||||
|
||||
def test_set_user_tier_downgrade(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test downgrading a user's tier from PRO to FREE."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_TARGET_EMAIL,
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
)
|
||||
mock_set = mocker.patch(
|
||||
f"{_MOCK_MODULE}.set_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "FREE"},
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert data["user_id"] == target_user_id
|
||||
assert data["tier"] == "FREE"
|
||||
mock_set.assert_awaited_once_with(target_user_id, SubscriptionTier.FREE)
|
||||
|
||||
|
||||
def test_set_user_tier_invalid_tier(
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that setting an invalid tier returns 422."""
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "invalid"},
|
||||
)
|
||||
|
||||
assert response.status_code == 422
|
||||
|
||||
|
||||
def test_set_user_tier_invalid_tier_uppercase(
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that setting an unrecognised uppercase tier (e.g. 'INVALID') returns 422.
|
||||
|
||||
Regression: ensures Pydantic enum validation rejects values that are not
|
||||
members of SubscriptionTier, even when they look like valid enum names.
|
||||
"""
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "INVALID"},
|
||||
)
|
||||
|
||||
assert response.status_code == 422
|
||||
body = response.json()
|
||||
assert "detail" in body
|
||||
|
||||
|
||||
def test_set_user_tier_email_lookup_failure_non_blocking(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""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"),
|
||||
)
|
||||
mocker.patch(
|
||||
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,
|
||||
)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "PRO"},
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
mock_set.assert_awaited_once()
|
||||
|
||||
|
||||
def test_set_user_tier_db_failure(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
target_user_id: str,
|
||||
) -> None:
|
||||
"""Test that DB failure on set tier returns 500."""
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_email_by_id",
|
||||
new_callable=AsyncMock,
|
||||
return_value=_TARGET_EMAIL,
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.FREE,
|
||||
)
|
||||
mocker.patch(
|
||||
f"{_MOCK_MODULE}.set_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
side_effect=Exception("DB connection refused"),
|
||||
)
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": target_user_id, "tier": "PRO"},
|
||||
)
|
||||
|
||||
assert response.status_code == 500
|
||||
|
||||
|
||||
def test_tier_endpoints_require_admin_role(mock_jwt_user) -> None:
|
||||
"""Test that tier admin endpoints require admin role."""
|
||||
app.dependency_overrides[get_jwt_payload] = mock_jwt_user["get_jwt_payload"]
|
||||
|
||||
response = client.get("/admin/rate_limit/tier", params={"user_id": "test"})
|
||||
assert response.status_code == 403
|
||||
|
||||
response = client.post(
|
||||
"/admin/rate_limit/tier",
|
||||
json={"user_id": "test", "tier": "PRO"},
|
||||
)
|
||||
assert response.status_code == 403
|
||||
|
||||
|
||||
# ─── search_users endpoint ──────────────────────────────────────────
|
||||
|
||||
|
||||
def test_search_users_returns_matching_users(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
admin_user_id: str,
|
||||
) -> None:
|
||||
"""Partial search should return all matching users from the User table."""
|
||||
mocker.patch(
|
||||
_MOCK_MODULE + ".search_users",
|
||||
new_callable=AsyncMock,
|
||||
return_value=[
|
||||
("user-1", "zamil.majdy@gmail.com"),
|
||||
("user-2", "zamil.majdy@agpt.co"),
|
||||
],
|
||||
)
|
||||
|
||||
response = client.get("/admin/rate_limit/search_users", params={"query": "zamil"})
|
||||
|
||||
assert response.status_code == 200
|
||||
results = response.json()
|
||||
assert len(results) == 2
|
||||
assert results[0]["user_email"] == "zamil.majdy@gmail.com"
|
||||
assert results[1]["user_email"] == "zamil.majdy@agpt.co"
|
||||
|
||||
|
||||
def test_search_users_empty_results(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
admin_user_id: str,
|
||||
) -> None:
|
||||
"""Search with no matches returns empty list."""
|
||||
mocker.patch(
|
||||
_MOCK_MODULE + ".search_users",
|
||||
new_callable=AsyncMock,
|
||||
return_value=[],
|
||||
)
|
||||
|
||||
response = client.get(
|
||||
"/admin/rate_limit/search_users", params={"query": "nonexistent"}
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.json() == []
|
||||
|
||||
|
||||
def test_search_users_short_query_rejected(
|
||||
admin_user_id: str,
|
||||
) -> None:
|
||||
"""Query shorter than 3 characters should return 400."""
|
||||
response = client.get("/admin/rate_limit/search_users", params={"query": "ab"})
|
||||
assert response.status_code == 400
|
||||
|
||||
|
||||
def test_search_users_negative_limit_clamped(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
admin_user_id: str,
|
||||
) -> None:
|
||||
"""Negative limit should be clamped to 1, not passed through."""
|
||||
mock_search = mocker.patch(
|
||||
_MOCK_MODULE + ".search_users",
|
||||
new_callable=AsyncMock,
|
||||
return_value=[],
|
||||
)
|
||||
|
||||
response = client.get(
|
||||
"/admin/rate_limit/search_users", params={"query": "test", "limit": -1}
|
||||
)
|
||||
|
||||
assert response.status_code == 200
|
||||
mock_search.assert_awaited_once_with("test", limit=1)
|
||||
|
||||
|
||||
def test_search_users_requires_admin_role(mock_jwt_user) -> None:
|
||||
"""Test that the search_users endpoint requires admin role."""
|
||||
app.dependency_overrides[get_jwt_payload] = mock_jwt_user["get_jwt_payload"]
|
||||
|
||||
response = client.get("/admin/rate_limit/search_users", params={"query": "test"})
|
||||
assert response.status_code == 403
|
||||
@@ -4,14 +4,14 @@ import asyncio
|
||||
import logging
|
||||
import re
|
||||
from collections.abc import AsyncGenerator
|
||||
from typing import Annotated, Literal
|
||||
from typing import Annotated
|
||||
from uuid import uuid4
|
||||
|
||||
from autogpt_libs import auth
|
||||
from fastapi import APIRouter, HTTPException, Query, Response, Security
|
||||
from fastapi.responses import StreamingResponse
|
||||
from prisma.models import UserWorkspaceFile
|
||||
from pydantic import BaseModel, ConfigDict, Field, field_validator
|
||||
from pydantic import BaseModel, Field, field_validator
|
||||
|
||||
from backend.copilot import service as chat_service
|
||||
from backend.copilot import stream_registry
|
||||
@@ -20,7 +20,6 @@ from backend.copilot.executor.utils import enqueue_cancel_task, enqueue_copilot_
|
||||
from backend.copilot.model import (
|
||||
ChatMessage,
|
||||
ChatSession,
|
||||
ChatSessionMetadata,
|
||||
append_and_save_message,
|
||||
create_chat_session,
|
||||
delete_chat_session,
|
||||
@@ -31,14 +30,8 @@ from backend.copilot.model import (
|
||||
from backend.copilot.rate_limit import (
|
||||
CoPilotUsageStatus,
|
||||
RateLimitExceeded,
|
||||
acquire_reset_lock,
|
||||
check_rate_limit,
|
||||
get_daily_reset_count,
|
||||
get_global_rate_limits,
|
||||
get_usage_status,
|
||||
increment_daily_reset_count,
|
||||
release_reset_lock,
|
||||
reset_daily_usage,
|
||||
)
|
||||
from backend.copilot.response_model import StreamError, StreamFinish, StreamHeartbeat
|
||||
from backend.copilot.tools.e2b_sandbox import kill_sandbox
|
||||
@@ -66,16 +59,9 @@ from backend.copilot.tools.models import (
|
||||
UnderstandingUpdatedResponse,
|
||||
)
|
||||
from backend.copilot.tracking import track_user_message
|
||||
from backend.data.credit import UsageTransactionMetadata, get_user_credit_model
|
||||
from backend.data.redis_client import get_redis_async
|
||||
from backend.data.understanding import get_business_understanding
|
||||
from backend.data.workspace import get_or_create_workspace
|
||||
from backend.util.exceptions import InsufficientBalanceError, NotFoundError
|
||||
from backend.util.settings import Settings
|
||||
|
||||
settings = Settings()
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
from backend.util.exceptions import NotFoundError
|
||||
|
||||
config = ChatConfig()
|
||||
|
||||
@@ -83,6 +69,8 @@ _UUID_RE = re.compile(
|
||||
r"^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$", re.I
|
||||
)
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
async def _validate_and_get_session(
|
||||
session_id: str,
|
||||
@@ -111,23 +99,6 @@ class StreamChatRequest(BaseModel):
|
||||
file_ids: list[str] | None = Field(
|
||||
default=None, max_length=20
|
||||
) # Workspace file IDs attached to this message
|
||||
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).",
|
||||
)
|
||||
|
||||
|
||||
class CreateSessionRequest(BaseModel):
|
||||
"""Request model for creating a new chat session.
|
||||
|
||||
``dry_run`` is a **top-level** field — do not nest it inside ``metadata``.
|
||||
Extra/unknown fields are rejected (422) to prevent silent mis-use.
|
||||
"""
|
||||
|
||||
model_config = ConfigDict(extra="forbid")
|
||||
|
||||
dry_run: bool = False
|
||||
|
||||
|
||||
class CreateSessionResponse(BaseModel):
|
||||
@@ -136,7 +107,6 @@ class CreateSessionResponse(BaseModel):
|
||||
id: str
|
||||
created_at: str
|
||||
user_id: str | None
|
||||
metadata: ChatSessionMetadata = ChatSessionMetadata()
|
||||
|
||||
|
||||
class ActiveStreamInfo(BaseModel):
|
||||
@@ -157,7 +127,6 @@ class SessionDetailResponse(BaseModel):
|
||||
active_stream: ActiveStreamInfo | None = None # Present if stream is still active
|
||||
total_prompt_tokens: int = 0
|
||||
total_completion_tokens: int = 0
|
||||
metadata: ChatSessionMetadata = ChatSessionMetadata()
|
||||
|
||||
|
||||
class SessionSummaryResponse(BaseModel):
|
||||
@@ -268,7 +237,6 @@ async def list_sessions(
|
||||
)
|
||||
async def create_session(
|
||||
user_id: Annotated[str, Security(auth.get_user_id)],
|
||||
request: CreateSessionRequest | None = None,
|
||||
) -> CreateSessionResponse:
|
||||
"""
|
||||
Create a new chat session.
|
||||
@@ -277,28 +245,22 @@ async def create_session(
|
||||
|
||||
Args:
|
||||
user_id: The authenticated user ID parsed from the JWT (required).
|
||||
request: Optional request body. When provided, ``dry_run=True``
|
||||
forces run_block and run_agent calls to use dry-run simulation.
|
||||
|
||||
Returns:
|
||||
CreateSessionResponse: Details of the created session.
|
||||
|
||||
"""
|
||||
dry_run = request.dry_run if request else False
|
||||
|
||||
logger.info(
|
||||
f"Creating session with user_id: "
|
||||
f"...{user_id[-8:] if len(user_id) > 8 else '<redacted>'}"
|
||||
f"{', dry_run=True' if dry_run else ''}"
|
||||
)
|
||||
|
||||
session = await create_chat_session(user_id, dry_run=dry_run)
|
||||
session = await create_chat_session(user_id)
|
||||
|
||||
return CreateSessionResponse(
|
||||
id=session.session_id,
|
||||
created_at=session.started_at.isoformat(),
|
||||
user_id=session.user_id,
|
||||
metadata=session.metadata,
|
||||
)
|
||||
|
||||
|
||||
@@ -447,7 +409,6 @@ async def get_session(
|
||||
active_stream=active_stream_info,
|
||||
total_prompt_tokens=total_prompt,
|
||||
total_completion_tokens=total_completion,
|
||||
metadata=session.metadata,
|
||||
)
|
||||
|
||||
|
||||
@@ -460,193 +421,11 @@ async def get_copilot_usage(
|
||||
"""Get CoPilot usage status for the authenticated user.
|
||||
|
||||
Returns current token usage vs limits for daily and weekly windows.
|
||||
Global defaults sourced from LaunchDarkly (falling back to config).
|
||||
Includes the user's rate-limit tier.
|
||||
"""
|
||||
daily_limit, weekly_limit, tier = await get_global_rate_limits(
|
||||
user_id, config.daily_token_limit, config.weekly_token_limit
|
||||
)
|
||||
return await get_usage_status(
|
||||
user_id=user_id,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
rate_limit_reset_cost=config.rate_limit_reset_cost,
|
||||
tier=tier,
|
||||
)
|
||||
|
||||
|
||||
class RateLimitResetResponse(BaseModel):
|
||||
"""Response from resetting the daily rate limit."""
|
||||
|
||||
success: bool
|
||||
credits_charged: int = Field(description="Credits charged (in cents)")
|
||||
remaining_balance: int = Field(description="Credit balance after charge (in cents)")
|
||||
usage: CoPilotUsageStatus = Field(description="Updated usage status after reset")
|
||||
|
||||
|
||||
@router.post(
|
||||
"/usage/reset",
|
||||
status_code=200,
|
||||
responses={
|
||||
400: {
|
||||
"description": "Bad Request (feature disabled or daily limit not reached)"
|
||||
},
|
||||
402: {"description": "Payment Required (insufficient credits)"},
|
||||
429: {
|
||||
"description": "Too Many Requests (max daily resets exceeded or reset in progress)"
|
||||
},
|
||||
503: {
|
||||
"description": "Service Unavailable (Redis reset failed; credits refunded or support needed)"
|
||||
},
|
||||
},
|
||||
)
|
||||
async def reset_copilot_usage(
|
||||
user_id: Annotated[str, Security(auth.get_user_id)],
|
||||
) -> RateLimitResetResponse:
|
||||
"""Reset the daily CoPilot rate limit by spending credits.
|
||||
|
||||
Allows users who have hit their daily token limit to spend credits
|
||||
to reset their daily usage counter and continue working.
|
||||
Returns 400 if the feature is disabled or the user is not over the limit.
|
||||
Returns 402 if the user has insufficient credits.
|
||||
"""
|
||||
cost = config.rate_limit_reset_cost
|
||||
if cost <= 0:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="Rate limit reset is not available.",
|
||||
)
|
||||
|
||||
if not settings.config.enable_credit:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="Rate limit reset is not available (credit system is disabled).",
|
||||
)
|
||||
|
||||
daily_limit, weekly_limit, tier = await get_global_rate_limits(
|
||||
user_id, config.daily_token_limit, config.weekly_token_limit
|
||||
)
|
||||
|
||||
if daily_limit <= 0:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="No daily limit is configured — nothing to reset.",
|
||||
)
|
||||
|
||||
# Check max daily resets. get_daily_reset_count returns None when Redis
|
||||
# is unavailable; reject the reset in that case to prevent unlimited
|
||||
# free resets when the counter store is down.
|
||||
reset_count = await get_daily_reset_count(user_id)
|
||||
if reset_count is None:
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail="Unable to verify reset eligibility — please try again later.",
|
||||
)
|
||||
if config.max_daily_resets > 0 and reset_count >= config.max_daily_resets:
|
||||
raise HTTPException(
|
||||
status_code=429,
|
||||
detail=f"You've used all {config.max_daily_resets} resets for today.",
|
||||
)
|
||||
|
||||
# Acquire a per-user lock to prevent TOCTOU races (concurrent resets).
|
||||
if not await acquire_reset_lock(user_id):
|
||||
raise HTTPException(
|
||||
status_code=429,
|
||||
detail="A reset is already in progress. Please try again.",
|
||||
)
|
||||
|
||||
try:
|
||||
# Verify the user is actually at or over their daily limit.
|
||||
# (rate_limit_reset_cost intentionally omitted — this object is only
|
||||
# used for limit checks, not returned to the client.)
|
||||
usage_status = await get_usage_status(
|
||||
user_id=user_id,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
tier=tier,
|
||||
)
|
||||
if daily_limit > 0 and usage_status.daily.used < daily_limit:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="You have not reached your daily limit yet.",
|
||||
)
|
||||
|
||||
# If the weekly limit is also exhausted, resetting the daily counter
|
||||
# won't help — the user would still be blocked by the weekly limit.
|
||||
if weekly_limit > 0 and usage_status.weekly.used >= weekly_limit:
|
||||
raise HTTPException(
|
||||
status_code=400,
|
||||
detail="Your weekly limit is also reached. Resetting the daily limit won't help.",
|
||||
)
|
||||
|
||||
# Charge credits.
|
||||
credit_model = await get_user_credit_model(user_id)
|
||||
try:
|
||||
remaining = await credit_model.spend_credits(
|
||||
user_id=user_id,
|
||||
cost=cost,
|
||||
metadata=UsageTransactionMetadata(
|
||||
reason="CoPilot daily rate limit reset",
|
||||
),
|
||||
)
|
||||
except InsufficientBalanceError as e:
|
||||
raise HTTPException(
|
||||
status_code=402,
|
||||
detail="Insufficient credits to reset your rate limit.",
|
||||
) from e
|
||||
|
||||
# Reset daily usage in Redis. If this fails, refund the credits
|
||||
# so the user is not charged for a service they did not receive.
|
||||
if not await reset_daily_usage(user_id, daily_token_limit=daily_limit):
|
||||
# Compensate: refund the charged credits.
|
||||
refunded = False
|
||||
try:
|
||||
await credit_model.top_up_credits(user_id, cost)
|
||||
refunded = True
|
||||
logger.warning(
|
||||
"Refunded %d credits to user %s after Redis reset failure",
|
||||
cost,
|
||||
user_id[:8],
|
||||
)
|
||||
except Exception:
|
||||
logger.error(
|
||||
"CRITICAL: Failed to refund %d credits to user %s "
|
||||
"after Redis reset failure — manual intervention required",
|
||||
cost,
|
||||
user_id[:8],
|
||||
exc_info=True,
|
||||
)
|
||||
if refunded:
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail="Rate limit reset failed — please try again later. "
|
||||
"Your credits have not been charged.",
|
||||
)
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail="Rate limit reset failed and the automatic refund "
|
||||
"also failed. Please contact support for assistance.",
|
||||
)
|
||||
|
||||
# Track the reset count for daily cap enforcement.
|
||||
await increment_daily_reset_count(user_id)
|
||||
finally:
|
||||
await release_reset_lock(user_id)
|
||||
|
||||
# Return updated usage status.
|
||||
updated_usage = await get_usage_status(
|
||||
user_id=user_id,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
rate_limit_reset_cost=config.rate_limit_reset_cost,
|
||||
tier=tier,
|
||||
)
|
||||
|
||||
return RateLimitResetResponse(
|
||||
success=True,
|
||||
credits_charged=cost,
|
||||
remaining_balance=remaining,
|
||||
usage=updated_usage,
|
||||
daily_token_limit=config.daily_token_limit,
|
||||
weekly_token_limit=config.weekly_token_limit,
|
||||
)
|
||||
|
||||
|
||||
@@ -747,16 +526,12 @@ async def stream_chat_post(
|
||||
|
||||
# Pre-turn rate limit check (token-based).
|
||||
# check_rate_limit short-circuits internally when both limits are 0.
|
||||
# Global defaults sourced from LaunchDarkly, falling back to config.
|
||||
if user_id:
|
||||
try:
|
||||
daily_limit, weekly_limit, _ = await get_global_rate_limits(
|
||||
user_id, config.daily_token_limit, config.weekly_token_limit
|
||||
)
|
||||
await check_rate_limit(
|
||||
user_id=user_id,
|
||||
daily_token_limit=daily_limit,
|
||||
weekly_token_limit=weekly_limit,
|
||||
daily_token_limit=config.daily_token_limit,
|
||||
weekly_token_limit=config.weekly_token_limit,
|
||||
)
|
||||
except RateLimitExceeded as e:
|
||||
raise HTTPException(status_code=429, detail=str(e)) from e
|
||||
@@ -845,7 +620,6 @@ async def stream_chat_post(
|
||||
is_user_message=request.is_user_message,
|
||||
context=request.context,
|
||||
file_ids=sanitized_file_ids,
|
||||
mode=request.mode,
|
||||
)
|
||||
|
||||
setup_time = (time.perf_counter() - stream_start_time) * 1000
|
||||
@@ -1120,47 +894,6 @@ async def session_assign_user(
|
||||
return {"status": "ok"}
|
||||
|
||||
|
||||
# ========== Suggested Prompts ==========
|
||||
|
||||
|
||||
class SuggestedTheme(BaseModel):
|
||||
"""A themed group of suggested prompts."""
|
||||
|
||||
name: str
|
||||
prompts: list[str]
|
||||
|
||||
|
||||
class SuggestedPromptsResponse(BaseModel):
|
||||
"""Response model for user-specific suggested prompts grouped by theme."""
|
||||
|
||||
themes: list[SuggestedTheme]
|
||||
|
||||
|
||||
@router.get(
|
||||
"/suggested-prompts",
|
||||
dependencies=[Security(auth.requires_user)],
|
||||
)
|
||||
async def get_suggested_prompts(
|
||||
user_id: Annotated[str, Security(auth.get_user_id)],
|
||||
) -> SuggestedPromptsResponse:
|
||||
"""
|
||||
Get LLM-generated suggested prompts grouped by theme.
|
||||
|
||||
Returns personalized quick-action prompts based on the user's
|
||||
business understanding. Returns empty themes list if no custom
|
||||
prompts are available.
|
||||
"""
|
||||
understanding = await get_business_understanding(user_id)
|
||||
if understanding is None or not understanding.suggested_prompts:
|
||||
return SuggestedPromptsResponse(themes=[])
|
||||
|
||||
themes = [
|
||||
SuggestedTheme(name=name, prompts=prompts)
|
||||
for name, prompts in understanding.suggested_prompts.items()
|
||||
]
|
||||
return SuggestedPromptsResponse(themes=themes)
|
||||
|
||||
|
||||
# ========== Configuration ==========
|
||||
|
||||
|
||||
@@ -1209,7 +942,7 @@ async def health_check() -> dict:
|
||||
)
|
||||
|
||||
# Create and retrieve session to verify full data layer
|
||||
session = await create_chat_session(health_check_user_id, dry_run=False)
|
||||
session = await create_chat_session(health_check_user_id)
|
||||
await get_chat_session(session.session_id, health_check_user_id)
|
||||
|
||||
return {
|
||||
|
||||
@@ -1,7 +1,7 @@
|
||||
"""Tests for chat API routes: session title update, file attachment validation, usage, and rate limiting."""
|
||||
|
||||
from datetime import UTC, datetime, timedelta
|
||||
from unittest.mock import AsyncMock, MagicMock
|
||||
from unittest.mock import AsyncMock
|
||||
|
||||
import fastapi
|
||||
import fastapi.testclient
|
||||
@@ -9,7 +9,6 @@ import pytest
|
||||
import pytest_mock
|
||||
|
||||
from backend.api.features.chat import routes as chat_routes
|
||||
from backend.copilot.rate_limit import SubscriptionTier
|
||||
|
||||
app = fastapi.FastAPI()
|
||||
app.include_router(chat_routes.router)
|
||||
@@ -332,28 +331,14 @@ def _mock_usage(
|
||||
*,
|
||||
daily_used: int = 500,
|
||||
weekly_used: int = 2000,
|
||||
daily_limit: int = 10000,
|
||||
weekly_limit: int = 50000,
|
||||
tier: "SubscriptionTier" = SubscriptionTier.FREE,
|
||||
) -> AsyncMock:
|
||||
"""Mock get_usage_status and get_global_rate_limits for usage endpoint tests.
|
||||
|
||||
Mocks both ``get_global_rate_limits`` (returns the given limits + tier) and
|
||||
``get_usage_status`` so that tests exercise the endpoint without hitting
|
||||
LaunchDarkly or Prisma.
|
||||
"""
|
||||
"""Mock get_usage_status to return a predictable CoPilotUsageStatus."""
|
||||
from backend.copilot.rate_limit import CoPilotUsageStatus, UsageWindow
|
||||
|
||||
mocker.patch(
|
||||
"backend.api.features.chat.routes.get_global_rate_limits",
|
||||
new_callable=AsyncMock,
|
||||
return_value=(daily_limit, weekly_limit, tier),
|
||||
)
|
||||
|
||||
resets_at = datetime.now(UTC) + timedelta(days=1)
|
||||
status = CoPilotUsageStatus(
|
||||
daily=UsageWindow(used=daily_used, limit=daily_limit, resets_at=resets_at),
|
||||
weekly=UsageWindow(used=weekly_used, limit=weekly_limit, resets_at=resets_at),
|
||||
daily=UsageWindow(used=daily_used, limit=10000, resets_at=resets_at),
|
||||
weekly=UsageWindow(used=weekly_used, limit=50000, resets_at=resets_at),
|
||||
)
|
||||
return mocker.patch(
|
||||
"backend.api.features.chat.routes.get_usage_status",
|
||||
@@ -383,8 +368,6 @@ def test_usage_returns_daily_and_weekly(
|
||||
user_id=test_user_id,
|
||||
daily_token_limit=10000,
|
||||
weekly_token_limit=50000,
|
||||
rate_limit_reset_cost=chat_routes.config.rate_limit_reset_cost,
|
||||
tier=SubscriptionTier.FREE,
|
||||
)
|
||||
|
||||
|
||||
@@ -392,10 +375,11 @@ def test_usage_uses_config_limits(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""The endpoint forwards resolved limits from get_global_rate_limits to get_usage_status."""
|
||||
mock_get = _mock_usage(mocker, daily_limit=99999, weekly_limit=77777)
|
||||
"""The endpoint forwards daily_token_limit and weekly_token_limit from config."""
|
||||
mock_get = _mock_usage(mocker)
|
||||
|
||||
mocker.patch.object(chat_routes.config, "rate_limit_reset_cost", 500)
|
||||
mocker.patch.object(chat_routes.config, "daily_token_limit", 99999)
|
||||
mocker.patch.object(chat_routes.config, "weekly_token_limit", 77777)
|
||||
|
||||
response = client.get("/usage")
|
||||
|
||||
@@ -404,8 +388,6 @@ def test_usage_uses_config_limits(
|
||||
user_id=test_user_id,
|
||||
daily_token_limit=99999,
|
||||
weekly_token_limit=77777,
|
||||
rate_limit_reset_cost=500,
|
||||
tier=SubscriptionTier.FREE,
|
||||
)
|
||||
|
||||
|
||||
@@ -418,126 +400,3 @@ def test_usage_rejects_unauthenticated_request() -> None:
|
||||
response = unauthenticated_client.get("/usage")
|
||||
|
||||
assert response.status_code == 401
|
||||
|
||||
|
||||
# ─── Suggested prompts endpoint ──────────────────────────────────────
|
||||
|
||||
|
||||
def _mock_get_business_understanding(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
*,
|
||||
return_value=None,
|
||||
):
|
||||
"""Mock get_business_understanding."""
|
||||
return mocker.patch(
|
||||
"backend.api.features.chat.routes.get_business_understanding",
|
||||
new_callable=AsyncMock,
|
||||
return_value=return_value,
|
||||
)
|
||||
|
||||
|
||||
def test_suggested_prompts_returns_themes(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""User with themed prompts gets them back as themes list."""
|
||||
mock_understanding = MagicMock()
|
||||
mock_understanding.suggested_prompts = {
|
||||
"Learn": ["L1", "L2"],
|
||||
"Create": ["C1"],
|
||||
}
|
||||
_mock_get_business_understanding(mocker, return_value=mock_understanding)
|
||||
|
||||
response = client.get("/suggested-prompts")
|
||||
|
||||
assert response.status_code == 200
|
||||
data = response.json()
|
||||
assert "themes" in data
|
||||
themes_by_name = {t["name"]: t["prompts"] for t in data["themes"]}
|
||||
assert themes_by_name["Learn"] == ["L1", "L2"]
|
||||
assert themes_by_name["Create"] == ["C1"]
|
||||
|
||||
|
||||
def test_suggested_prompts_no_understanding(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""User with no understanding gets empty themes list."""
|
||||
_mock_get_business_understanding(mocker, return_value=None)
|
||||
|
||||
response = client.get("/suggested-prompts")
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.json() == {"themes": []}
|
||||
|
||||
|
||||
def test_suggested_prompts_empty_prompts(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""User with understanding but empty prompts gets empty themes list."""
|
||||
mock_understanding = MagicMock()
|
||||
mock_understanding.suggested_prompts = {}
|
||||
_mock_get_business_understanding(mocker, return_value=mock_understanding)
|
||||
|
||||
response = client.get("/suggested-prompts")
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.json() == {"themes": []}
|
||||
|
||||
|
||||
# ─── Create session: dry_run contract ─────────────────────────────────
|
||||
|
||||
|
||||
def _mock_create_chat_session(mocker: pytest_mock.MockerFixture):
|
||||
"""Mock create_chat_session to return a fake session."""
|
||||
from backend.copilot.model import ChatSession
|
||||
|
||||
async def _fake_create(user_id: str, *, dry_run: bool):
|
||||
return ChatSession.new(user_id, dry_run=dry_run)
|
||||
|
||||
return mocker.patch(
|
||||
"backend.api.features.chat.routes.create_chat_session",
|
||||
new_callable=AsyncMock,
|
||||
side_effect=_fake_create,
|
||||
)
|
||||
|
||||
|
||||
def test_create_session_dry_run_true(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""Sending ``{"dry_run": true}`` sets metadata.dry_run to True."""
|
||||
_mock_create_chat_session(mocker)
|
||||
|
||||
response = client.post("/sessions", json={"dry_run": True})
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.json()["metadata"]["dry_run"] is True
|
||||
|
||||
|
||||
def test_create_session_dry_run_default_false(
|
||||
mocker: pytest_mock.MockerFixture,
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""Empty body defaults dry_run to False."""
|
||||
_mock_create_chat_session(mocker)
|
||||
|
||||
response = client.post("/sessions")
|
||||
|
||||
assert response.status_code == 200
|
||||
assert response.json()["metadata"]["dry_run"] is False
|
||||
|
||||
|
||||
def test_create_session_rejects_nested_metadata(
|
||||
test_user_id: str,
|
||||
) -> None:
|
||||
"""Sending ``{"metadata": {"dry_run": true}}`` must return 422, not silently
|
||||
default to ``dry_run=False``. This guards against the common mistake of
|
||||
nesting dry_run inside metadata instead of providing it at the top level."""
|
||||
response = client.post(
|
||||
"/sessions",
|
||||
json={"metadata": {"dry_run": True}},
|
||||
)
|
||||
|
||||
assert response.status_code == 422
|
||||
|
||||
@@ -40,15 +40,11 @@ from backend.data.onboarding import OnboardingStep, complete_onboarding_step
|
||||
from backend.data.user import get_user_integrations
|
||||
from backend.executor.utils import add_graph_execution
|
||||
from backend.integrations.ayrshare import AyrshareClient, SocialPlatform
|
||||
from backend.integrations.credentials_store import (
|
||||
is_system_credential,
|
||||
provider_matches,
|
||||
)
|
||||
from backend.integrations.credentials_store import provider_matches
|
||||
from backend.integrations.creds_manager import (
|
||||
IntegrationCredentialsManager,
|
||||
create_mcp_oauth_handler,
|
||||
)
|
||||
from backend.integrations.managed_credentials import ensure_managed_credentials
|
||||
from backend.integrations.oauth import CREDENTIALS_BY_PROVIDER, HANDLERS_BY_NAME
|
||||
from backend.integrations.providers import ProviderName
|
||||
from backend.integrations.webhooks import get_webhook_manager
|
||||
@@ -114,7 +110,6 @@ class CredentialsMetaResponse(BaseModel):
|
||||
default=None,
|
||||
description="Host pattern for host-scoped or MCP server URL for MCP credentials",
|
||||
)
|
||||
is_managed: bool = False
|
||||
|
||||
@model_validator(mode="before")
|
||||
@classmethod
|
||||
@@ -153,7 +148,6 @@ def to_meta_response(cred: Credentials) -> CredentialsMetaResponse:
|
||||
scopes=cred.scopes if isinstance(cred, OAuth2Credentials) else None,
|
||||
username=cred.username if isinstance(cred, OAuth2Credentials) else None,
|
||||
host=CredentialsMetaResponse.get_host(cred),
|
||||
is_managed=cred.is_managed,
|
||||
)
|
||||
|
||||
|
||||
@@ -230,9 +224,6 @@ async def callback(
|
||||
async def list_credentials(
|
||||
user_id: Annotated[str, Security(get_user_id)],
|
||||
) -> list[CredentialsMetaResponse]:
|
||||
# Fire-and-forget: provision missing managed credentials in the background.
|
||||
# The credential appears on the next page load; listing is never blocked.
|
||||
asyncio.create_task(ensure_managed_credentials(user_id, creds_manager.store))
|
||||
credentials = await creds_manager.store.get_all_creds(user_id)
|
||||
|
||||
return [
|
||||
@@ -247,7 +238,6 @@ async def list_credentials_by_provider(
|
||||
],
|
||||
user_id: Annotated[str, Security(get_user_id)],
|
||||
) -> list[CredentialsMetaResponse]:
|
||||
asyncio.create_task(ensure_managed_credentials(user_id, creds_manager.store))
|
||||
credentials = await creds_manager.store.get_creds_by_provider(user_id, provider)
|
||||
|
||||
return [
|
||||
@@ -342,11 +332,6 @@ async def delete_credentials(
|
||||
raise HTTPException(
|
||||
status_code=status.HTTP_404_NOT_FOUND, detail="Credentials not found"
|
||||
)
|
||||
if is_system_credential(cred_id):
|
||||
raise HTTPException(
|
||||
status_code=status.HTTP_403_FORBIDDEN,
|
||||
detail="System-managed credentials cannot be deleted",
|
||||
)
|
||||
creds = await creds_manager.store.get_creds_by_id(user_id, cred_id)
|
||||
if not creds:
|
||||
raise HTTPException(
|
||||
@@ -357,11 +342,6 @@ async def delete_credentials(
|
||||
status_code=status.HTTP_404_NOT_FOUND,
|
||||
detail="Credentials not found",
|
||||
)
|
||||
if creds.is_managed:
|
||||
raise HTTPException(
|
||||
status_code=status.HTTP_403_FORBIDDEN,
|
||||
detail="AutoGPT-managed credentials cannot be deleted",
|
||||
)
|
||||
|
||||
try:
|
||||
await remove_all_webhooks_for_credentials(user_id, creds, force)
|
||||
|
||||
@@ -1,7 +1,6 @@
|
||||
"""Tests for credentials API security: no secret leakage, SDK defaults filtered."""
|
||||
|
||||
from contextlib import asynccontextmanager
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
from unittest.mock import AsyncMock, patch
|
||||
|
||||
import fastapi
|
||||
import fastapi.testclient
|
||||
@@ -277,294 +276,3 @@ class TestCreateCredentialNoSecretInResponse:
|
||||
|
||||
assert resp.status_code == 403
|
||||
mock_mgr.create.assert_not_called()
|
||||
|
||||
|
||||
class TestManagedCredentials:
|
||||
"""AutoGPT-managed credentials cannot be deleted by users."""
|
||||
|
||||
def test_delete_is_managed_returns_403(self):
|
||||
cred = APIKeyCredentials(
|
||||
id="managed-cred-1",
|
||||
provider="agent_mail",
|
||||
title="AgentMail (managed by AutoGPT)",
|
||||
api_key=SecretStr("sk-managed-key"),
|
||||
is_managed=True,
|
||||
)
|
||||
with patch(
|
||||
"backend.api.features.integrations.router.creds_manager"
|
||||
) as mock_mgr:
|
||||
mock_mgr.store.get_creds_by_id = AsyncMock(return_value=cred)
|
||||
resp = client.request("DELETE", "/agent_mail/credentials/managed-cred-1")
|
||||
|
||||
assert resp.status_code == 403
|
||||
assert "AutoGPT-managed" in resp.json()["detail"]
|
||||
|
||||
def test_list_credentials_includes_is_managed_field(self):
|
||||
managed = APIKeyCredentials(
|
||||
id="managed-1",
|
||||
provider="agent_mail",
|
||||
title="AgentMail (managed)",
|
||||
api_key=SecretStr("sk-key"),
|
||||
is_managed=True,
|
||||
)
|
||||
regular = APIKeyCredentials(
|
||||
id="regular-1",
|
||||
provider="openai",
|
||||
title="My Key",
|
||||
api_key=SecretStr("sk-key"),
|
||||
)
|
||||
with patch(
|
||||
"backend.api.features.integrations.router.creds_manager"
|
||||
) as mock_mgr:
|
||||
mock_mgr.store.get_all_creds = AsyncMock(return_value=[managed, regular])
|
||||
resp = client.get("/credentials")
|
||||
|
||||
assert resp.status_code == 200
|
||||
data = resp.json()
|
||||
managed_cred = next(c for c in data if c["id"] == "managed-1")
|
||||
regular_cred = next(c for c in data if c["id"] == "regular-1")
|
||||
assert managed_cred["is_managed"] is True
|
||||
assert regular_cred["is_managed"] is False
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Managed credential provisioning infrastructure
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
def _make_managed_cred(
|
||||
provider: str = "agent_mail", pod_id: str = "pod-abc"
|
||||
) -> APIKeyCredentials:
|
||||
return APIKeyCredentials(
|
||||
id="managed-auto",
|
||||
provider=provider,
|
||||
title="AgentMail (managed by AutoGPT)",
|
||||
api_key=SecretStr("sk-pod-key"),
|
||||
is_managed=True,
|
||||
metadata={"pod_id": pod_id},
|
||||
)
|
||||
|
||||
|
||||
def _make_store_mock(**kwargs) -> MagicMock:
|
||||
"""Create a store mock with a working async ``locks()`` context manager."""
|
||||
|
||||
@asynccontextmanager
|
||||
async def _noop_locked(key):
|
||||
yield
|
||||
|
||||
locks_obj = MagicMock()
|
||||
locks_obj.locked = _noop_locked
|
||||
|
||||
store = MagicMock(**kwargs)
|
||||
store.locks = AsyncMock(return_value=locks_obj)
|
||||
return store
|
||||
|
||||
|
||||
class TestEnsureManagedCredentials:
|
||||
"""Unit tests for the ensure/cleanup helpers in managed_credentials.py."""
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_provisions_when_missing(self):
|
||||
"""Provider.provision() is called when no managed credential exists."""
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
_provisioned_users,
|
||||
ensure_managed_credentials,
|
||||
)
|
||||
|
||||
cred = _make_managed_cred()
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "test_provider"
|
||||
provider.is_available = AsyncMock(return_value=True)
|
||||
provider.provision = AsyncMock(return_value=cred)
|
||||
|
||||
store = _make_store_mock()
|
||||
store.has_managed_credential = AsyncMock(return_value=False)
|
||||
store.add_managed_credential = AsyncMock()
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["test_provider"] = provider
|
||||
_provisioned_users.pop("user-1", None)
|
||||
try:
|
||||
await ensure_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
_provisioned_users.pop("user-1", None)
|
||||
|
||||
provider.provision.assert_awaited_once_with("user-1")
|
||||
store.add_managed_credential.assert_awaited_once_with("user-1", cred)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skips_when_already_exists(self):
|
||||
"""Provider.provision() is NOT called when managed credential exists."""
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
_provisioned_users,
|
||||
ensure_managed_credentials,
|
||||
)
|
||||
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "test_provider"
|
||||
provider.is_available = AsyncMock(return_value=True)
|
||||
provider.provision = AsyncMock()
|
||||
|
||||
store = _make_store_mock()
|
||||
store.has_managed_credential = AsyncMock(return_value=True)
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["test_provider"] = provider
|
||||
_provisioned_users.pop("user-1", None)
|
||||
try:
|
||||
await ensure_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
_provisioned_users.pop("user-1", None)
|
||||
|
||||
provider.provision.assert_not_awaited()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skips_when_unavailable(self):
|
||||
"""Provider.provision() is NOT called when provider is not available."""
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
_provisioned_users,
|
||||
ensure_managed_credentials,
|
||||
)
|
||||
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "test_provider"
|
||||
provider.is_available = AsyncMock(return_value=False)
|
||||
provider.provision = AsyncMock()
|
||||
|
||||
store = _make_store_mock()
|
||||
store.has_managed_credential = AsyncMock()
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["test_provider"] = provider
|
||||
_provisioned_users.pop("user-1", None)
|
||||
try:
|
||||
await ensure_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
_provisioned_users.pop("user-1", None)
|
||||
|
||||
provider.provision.assert_not_awaited()
|
||||
store.has_managed_credential.assert_not_awaited()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_provision_failure_does_not_propagate(self):
|
||||
"""A failed provision is logged but does not raise."""
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
_provisioned_users,
|
||||
ensure_managed_credentials,
|
||||
)
|
||||
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "test_provider"
|
||||
provider.is_available = AsyncMock(return_value=True)
|
||||
provider.provision = AsyncMock(side_effect=RuntimeError("boom"))
|
||||
|
||||
store = _make_store_mock()
|
||||
store.has_managed_credential = AsyncMock(return_value=False)
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["test_provider"] = provider
|
||||
_provisioned_users.pop("user-1", None)
|
||||
try:
|
||||
await ensure_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
_provisioned_users.pop("user-1", None)
|
||||
|
||||
# No exception raised — provisioning failure is swallowed.
|
||||
|
||||
|
||||
class TestCleanupManagedCredentials:
|
||||
"""Unit tests for cleanup_managed_credentials."""
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_calls_deprovision_for_managed_creds(self):
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
cleanup_managed_credentials,
|
||||
)
|
||||
|
||||
cred = _make_managed_cred()
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "agent_mail"
|
||||
provider.deprovision = AsyncMock()
|
||||
|
||||
store = MagicMock()
|
||||
store.get_all_creds = AsyncMock(return_value=[cred])
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["agent_mail"] = provider
|
||||
try:
|
||||
await cleanup_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
|
||||
provider.deprovision.assert_awaited_once_with("user-1", cred)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skips_non_managed_creds(self):
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
cleanup_managed_credentials,
|
||||
)
|
||||
|
||||
regular = _make_api_key_cred()
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "openai"
|
||||
provider.deprovision = AsyncMock()
|
||||
|
||||
store = MagicMock()
|
||||
store.get_all_creds = AsyncMock(return_value=[regular])
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["openai"] = provider
|
||||
try:
|
||||
await cleanup_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
|
||||
provider.deprovision.assert_not_awaited()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_deprovision_failure_does_not_propagate(self):
|
||||
from backend.integrations.managed_credentials import (
|
||||
_PROVIDERS,
|
||||
cleanup_managed_credentials,
|
||||
)
|
||||
|
||||
cred = _make_managed_cred()
|
||||
provider = MagicMock()
|
||||
provider.provider_name = "agent_mail"
|
||||
provider.deprovision = AsyncMock(side_effect=RuntimeError("boom"))
|
||||
|
||||
store = MagicMock()
|
||||
store.get_all_creds = AsyncMock(return_value=[cred])
|
||||
|
||||
saved = dict(_PROVIDERS)
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS["agent_mail"] = provider
|
||||
try:
|
||||
await cleanup_managed_credentials("user-1", store)
|
||||
finally:
|
||||
_PROVIDERS.clear()
|
||||
_PROVIDERS.update(saved)
|
||||
|
||||
# No exception raised — cleanup failure is swallowed.
|
||||
|
||||
@@ -17,6 +17,8 @@ from backend.data.includes import library_agent_include
|
||||
from backend.util.exceptions import NotFoundError
|
||||
from backend.util.json import SafeJson
|
||||
|
||||
from .db import get_library_agent_by_graph_id, update_library_agent
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
@@ -59,17 +61,28 @@ async def add_graph_to_library(
|
||||
graph_model: GraphModel,
|
||||
user_id: str,
|
||||
) -> library_model.LibraryAgent:
|
||||
"""Check existing / restore soft-deleted / create new LibraryAgent.
|
||||
"""Check existing / restore soft-deleted / create new LibraryAgent."""
|
||||
if existing := await get_library_agent_by_graph_id(
|
||||
user_id, graph_model.id, graph_model.version
|
||||
):
|
||||
return existing
|
||||
|
||||
Uses a create-then-catch-UniqueViolationError-then-update pattern on
|
||||
the (userId, agentGraphId, agentGraphVersion) composite unique constraint.
|
||||
This is more robust than ``upsert`` because Prisma's upsert atomicity
|
||||
guarantees are not well-documented for all versions.
|
||||
"""
|
||||
settings_json = SafeJson(GraphSettings.from_graph(graph_model).model_dump())
|
||||
_include = library_agent_include(
|
||||
user_id, include_nodes=False, include_executions=False
|
||||
deleted_agent = await prisma.models.LibraryAgent.prisma().find_unique(
|
||||
where={
|
||||
"userId_agentGraphId_agentGraphVersion": {
|
||||
"userId": user_id,
|
||||
"agentGraphId": graph_model.id,
|
||||
"agentGraphVersion": graph_model.version,
|
||||
}
|
||||
},
|
||||
)
|
||||
if deleted_agent and (deleted_agent.isDeleted or deleted_agent.isArchived):
|
||||
return await update_library_agent(
|
||||
deleted_agent.id,
|
||||
user_id,
|
||||
is_deleted=False,
|
||||
is_archived=False,
|
||||
)
|
||||
|
||||
try:
|
||||
added_agent = await prisma.models.LibraryAgent.prisma().create(
|
||||
@@ -85,32 +98,23 @@ async def add_graph_to_library(
|
||||
},
|
||||
"isCreatedByUser": False,
|
||||
"useGraphIsActiveVersion": False,
|
||||
"settings": settings_json,
|
||||
"settings": SafeJson(
|
||||
GraphSettings.from_graph(graph_model).model_dump()
|
||||
),
|
||||
},
|
||||
include=_include,
|
||||
include=library_agent_include(
|
||||
user_id, include_nodes=False, include_executions=False
|
||||
),
|
||||
)
|
||||
except prisma.errors.UniqueViolationError:
|
||||
# Already exists — update to restore if previously soft-deleted/archived
|
||||
added_agent = await prisma.models.LibraryAgent.prisma().update(
|
||||
where={
|
||||
"userId_agentGraphId_agentGraphVersion": {
|
||||
"userId": user_id,
|
||||
"agentGraphId": graph_model.id,
|
||||
"agentGraphVersion": graph_model.version,
|
||||
}
|
||||
},
|
||||
data={
|
||||
"isDeleted": False,
|
||||
"isArchived": False,
|
||||
"settings": settings_json,
|
||||
},
|
||||
include=_include,
|
||||
# Race condition: concurrent request created the row between our
|
||||
# check and create. Re-read instead of crashing.
|
||||
existing = await get_library_agent_by_graph_id(
|
||||
user_id, graph_model.id, graph_model.version
|
||||
)
|
||||
if added_agent is None:
|
||||
raise NotFoundError(
|
||||
f"LibraryAgent for graph #{graph_model.id} "
|
||||
f"v{graph_model.version} not found after UniqueViolationError"
|
||||
)
|
||||
if existing:
|
||||
return existing
|
||||
raise # Shouldn't happen, but don't swallow unexpected errors
|
||||
|
||||
logger.debug(
|
||||
f"Added graph #{graph_model.id} v{graph_model.version} "
|
||||
|
||||
@@ -1,80 +1,71 @@
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
|
||||
import prisma.errors
|
||||
import pytest
|
||||
|
||||
from ._add_to_library import add_graph_to_library
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_add_graph_to_library_create_new_agent() -> None:
|
||||
"""When no matching LibraryAgent exists, create inserts a new one."""
|
||||
graph_model = MagicMock(id="graph-id", version=2, nodes=[])
|
||||
created_agent = MagicMock(name="CreatedLibraryAgent")
|
||||
converted_agent = MagicMock(name="ConvertedLibraryAgent")
|
||||
async def test_add_graph_to_library_restores_archived_agent() -> None:
|
||||
graph_model = MagicMock(id="graph-id", version=2)
|
||||
archived_agent = MagicMock(id="library-agent-id", isDeleted=False, isArchived=True)
|
||||
restored_agent = MagicMock(name="LibraryAgentModel")
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.get_library_agent_by_graph_id",
|
||||
new=AsyncMock(return_value=None),
|
||||
),
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.prisma.models.LibraryAgent.prisma"
|
||||
) as mock_prisma,
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.library_model.LibraryAgent.from_db",
|
||||
return_value=converted_agent,
|
||||
) as mock_from_db,
|
||||
"backend.api.features.library._add_to_library.update_library_agent",
|
||||
new=AsyncMock(return_value=restored_agent),
|
||||
) as mock_update,
|
||||
):
|
||||
mock_prisma.return_value.create = AsyncMock(return_value=created_agent)
|
||||
mock_prisma.return_value.find_unique = AsyncMock(return_value=archived_agent)
|
||||
|
||||
result = await add_graph_to_library("slv-id", graph_model, "user-id")
|
||||
|
||||
assert result is converted_agent
|
||||
mock_from_db.assert_called_once_with(created_agent)
|
||||
# Verify create was called with correct data
|
||||
create_call = mock_prisma.return_value.create.call_args
|
||||
create_data = create_call.kwargs["data"]
|
||||
assert create_data["User"] == {"connect": {"id": "user-id"}}
|
||||
assert create_data["AgentGraph"] == {
|
||||
"connect": {"graphVersionId": {"id": "graph-id", "version": 2}}
|
||||
}
|
||||
assert create_data["isCreatedByUser"] is False
|
||||
assert create_data["useGraphIsActiveVersion"] is False
|
||||
assert result is restored_agent
|
||||
mock_update.assert_awaited_once_with(
|
||||
"library-agent-id",
|
||||
"user-id",
|
||||
is_deleted=False,
|
||||
is_archived=False,
|
||||
)
|
||||
mock_prisma.return_value.create.assert_not_called()
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_add_graph_to_library_unique_violation_updates_existing() -> None:
|
||||
"""UniqueViolationError on create falls back to update."""
|
||||
graph_model = MagicMock(id="graph-id", version=2, nodes=[])
|
||||
updated_agent = MagicMock(name="UpdatedLibraryAgent")
|
||||
converted_agent = MagicMock(name="ConvertedLibraryAgent")
|
||||
async def test_add_graph_to_library_restores_deleted_agent() -> None:
|
||||
graph_model = MagicMock(id="graph-id", version=2)
|
||||
deleted_agent = MagicMock(id="library-agent-id", isDeleted=True, isArchived=False)
|
||||
restored_agent = MagicMock(name="LibraryAgentModel")
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.get_library_agent_by_graph_id",
|
||||
new=AsyncMock(return_value=None),
|
||||
),
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.prisma.models.LibraryAgent.prisma"
|
||||
) as mock_prisma,
|
||||
patch(
|
||||
"backend.api.features.library._add_to_library.library_model.LibraryAgent.from_db",
|
||||
return_value=converted_agent,
|
||||
) as mock_from_db,
|
||||
"backend.api.features.library._add_to_library.update_library_agent",
|
||||
new=AsyncMock(return_value=restored_agent),
|
||||
) as mock_update,
|
||||
):
|
||||
mock_prisma.return_value.create = AsyncMock(
|
||||
side_effect=prisma.errors.UniqueViolationError(
|
||||
MagicMock(), message="unique constraint"
|
||||
)
|
||||
)
|
||||
mock_prisma.return_value.update = AsyncMock(return_value=updated_agent)
|
||||
mock_prisma.return_value.find_unique = AsyncMock(return_value=deleted_agent)
|
||||
|
||||
result = await add_graph_to_library("slv-id", graph_model, "user-id")
|
||||
|
||||
assert result is converted_agent
|
||||
mock_from_db.assert_called_once_with(updated_agent)
|
||||
# Verify update was called with correct where and data
|
||||
update_call = mock_prisma.return_value.update.call_args
|
||||
assert update_call.kwargs["where"] == {
|
||||
"userId_agentGraphId_agentGraphVersion": {
|
||||
"userId": "user-id",
|
||||
"agentGraphId": "graph-id",
|
||||
"agentGraphVersion": 2,
|
||||
}
|
||||
}
|
||||
update_data = update_call.kwargs["data"]
|
||||
assert update_data["isDeleted"] is False
|
||||
assert update_data["isArchived"] is False
|
||||
assert result is restored_agent
|
||||
mock_update.assert_awaited_once_with(
|
||||
"library-agent-id",
|
||||
"user-id",
|
||||
is_deleted=False,
|
||||
is_archived=False,
|
||||
)
|
||||
mock_prisma.return_value.create.assert_not_called()
|
||||
|
||||
@@ -436,58 +436,32 @@ async def create_library_agent(
|
||||
async with transaction() as tx:
|
||||
library_agents = await asyncio.gather(
|
||||
*(
|
||||
prisma.models.LibraryAgent.prisma(tx).upsert(
|
||||
where={
|
||||
"userId_agentGraphId_agentGraphVersion": {
|
||||
"userId": user_id,
|
||||
"agentGraphId": graph_entry.id,
|
||||
"agentGraphVersion": graph_entry.version,
|
||||
}
|
||||
},
|
||||
data={
|
||||
"create": prisma.types.LibraryAgentCreateInput(
|
||||
isCreatedByUser=(user_id == graph.user_id),
|
||||
useGraphIsActiveVersion=True,
|
||||
User={"connect": {"id": user_id}},
|
||||
AgentGraph={
|
||||
"connect": {
|
||||
"graphVersionId": {
|
||||
"id": graph_entry.id,
|
||||
"version": graph_entry.version,
|
||||
}
|
||||
prisma.models.LibraryAgent.prisma(tx).create(
|
||||
data=prisma.types.LibraryAgentCreateInput(
|
||||
isCreatedByUser=(user_id == user_id),
|
||||
useGraphIsActiveVersion=True,
|
||||
User={"connect": {"id": user_id}},
|
||||
AgentGraph={
|
||||
"connect": {
|
||||
"graphVersionId": {
|
||||
"id": graph_entry.id,
|
||||
"version": graph_entry.version,
|
||||
}
|
||||
},
|
||||
settings=SafeJson(
|
||||
GraphSettings.from_graph(
|
||||
graph_entry,
|
||||
hitl_safe_mode=hitl_safe_mode,
|
||||
sensitive_action_safe_mode=sensitive_action_safe_mode,
|
||||
).model_dump()
|
||||
),
|
||||
**(
|
||||
{"Folder": {"connect": {"id": folder_id}}}
|
||||
if folder_id and graph_entry is graph
|
||||
else {}
|
||||
),
|
||||
),
|
||||
"update": {
|
||||
"isDeleted": False,
|
||||
"isArchived": False,
|
||||
"useGraphIsActiveVersion": True,
|
||||
"settings": SafeJson(
|
||||
GraphSettings.from_graph(
|
||||
graph_entry,
|
||||
hitl_safe_mode=hitl_safe_mode,
|
||||
sensitive_action_safe_mode=sensitive_action_safe_mode,
|
||||
).model_dump()
|
||||
),
|
||||
**(
|
||||
{"Folder": {"connect": {"id": folder_id}}}
|
||||
if folder_id and graph_entry is graph
|
||||
else {}
|
||||
),
|
||||
}
|
||||
},
|
||||
},
|
||||
settings=SafeJson(
|
||||
GraphSettings.from_graph(
|
||||
graph_entry,
|
||||
hitl_safe_mode=hitl_safe_mode,
|
||||
sensitive_action_safe_mode=sensitive_action_safe_mode,
|
||||
).model_dump()
|
||||
),
|
||||
**(
|
||||
{"Folder": {"connect": {"id": folder_id}}}
|
||||
if folder_id and graph_entry is graph
|
||||
else {}
|
||||
),
|
||||
),
|
||||
include=library_agent_include(
|
||||
user_id, include_nodes=False, include_executions=False
|
||||
),
|
||||
|
||||
@@ -1,6 +1,4 @@
|
||||
from contextlib import asynccontextmanager
|
||||
from datetime import datetime
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
|
||||
import prisma.enums
|
||||
import prisma.models
|
||||
@@ -87,6 +85,10 @@ async def test_get_library_agents(mocker):
|
||||
async def test_add_agent_to_library(mocker):
|
||||
await connect()
|
||||
|
||||
# Mock the transaction context
|
||||
mock_transaction = mocker.patch("backend.api.features.library.db.transaction")
|
||||
mock_transaction.return_value.__aenter__ = mocker.AsyncMock(return_value=None)
|
||||
mock_transaction.return_value.__aexit__ = mocker.AsyncMock(return_value=None)
|
||||
# Mock data
|
||||
mock_store_listing_data = prisma.models.StoreListingVersion(
|
||||
id="version123",
|
||||
@@ -141,11 +143,13 @@ async def test_add_agent_to_library(mocker):
|
||||
)
|
||||
|
||||
mock_library_agent = mocker.patch("prisma.models.LibraryAgent.prisma")
|
||||
mock_library_agent.return_value.find_first = mocker.AsyncMock(return_value=None)
|
||||
mock_library_agent.return_value.find_unique = mocker.AsyncMock(return_value=None)
|
||||
mock_library_agent.return_value.create = mocker.AsyncMock(
|
||||
return_value=mock_library_agent_data
|
||||
)
|
||||
|
||||
# Mock graph_db.get_graph function that's called in resolve_graph_for_library
|
||||
# Mock graph_db.get_graph function that's called to check for HITL blocks
|
||||
# (lives in _add_to_library.py after refactor, not db.py)
|
||||
mock_graph_db = mocker.patch(
|
||||
"backend.api.features.library._add_to_library.graph_db"
|
||||
@@ -171,27 +175,37 @@ async def test_add_agent_to_library(mocker):
|
||||
mock_store_listing_version.return_value.find_unique.assert_called_once_with(
|
||||
where={"id": "version123"}, include={"AgentGraph": True}
|
||||
)
|
||||
mock_library_agent.return_value.find_unique.assert_called_once_with(
|
||||
where={
|
||||
"userId_agentGraphId_agentGraphVersion": {
|
||||
"userId": "test-user",
|
||||
"agentGraphId": "agent1",
|
||||
"agentGraphVersion": 1,
|
||||
}
|
||||
},
|
||||
)
|
||||
# Check that create was called with the expected data including settings
|
||||
create_call_args = mock_library_agent.return_value.create.call_args
|
||||
assert create_call_args is not None
|
||||
|
||||
# Verify the create data structure
|
||||
create_data = create_call_args.kwargs["data"]
|
||||
expected_create = {
|
||||
# Verify the main structure
|
||||
expected_data = {
|
||||
"User": {"connect": {"id": "test-user"}},
|
||||
"AgentGraph": {"connect": {"graphVersionId": {"id": "agent1", "version": 1}}},
|
||||
"isCreatedByUser": False,
|
||||
"useGraphIsActiveVersion": False,
|
||||
}
|
||||
for key, value in expected_create.items():
|
||||
assert create_data[key] == value
|
||||
|
||||
actual_data = create_call_args[1]["data"]
|
||||
# Check that all expected fields are present
|
||||
for key, value in expected_data.items():
|
||||
assert actual_data[key] == value
|
||||
|
||||
# Check that settings field is present and is a SafeJson object
|
||||
assert "settings" in create_data
|
||||
assert hasattr(create_data["settings"], "__class__") # Should be a SafeJson object
|
||||
assert "settings" in actual_data
|
||||
assert hasattr(actual_data["settings"], "__class__") # Should be a SafeJson object
|
||||
|
||||
# Check include parameter
|
||||
assert create_call_args.kwargs["include"] == library_agent_include(
|
||||
assert create_call_args[1]["include"] == library_agent_include(
|
||||
"test-user", include_nodes=False, include_executions=False
|
||||
)
|
||||
|
||||
@@ -306,50 +320,3 @@ async def test_update_graph_in_library_allows_archived_library_agent(mocker):
|
||||
include_archived=True,
|
||||
)
|
||||
mock_update_library_agent.assert_awaited_once_with("test-user", created_graph)
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_create_library_agent_uses_upsert():
|
||||
"""create_library_agent should use upsert (not create) to handle duplicates."""
|
||||
mock_graph = MagicMock()
|
||||
mock_graph.id = "graph-1"
|
||||
mock_graph.version = 1
|
||||
mock_graph.user_id = "user-1"
|
||||
mock_graph.nodes = []
|
||||
mock_graph.sub_graphs = []
|
||||
|
||||
mock_upserted = MagicMock(name="UpsertedLibraryAgent")
|
||||
|
||||
@asynccontextmanager
|
||||
async def fake_tx():
|
||||
yield None
|
||||
|
||||
with (
|
||||
patch("backend.api.features.library.db.transaction", fake_tx),
|
||||
patch("prisma.models.LibraryAgent.prisma") as mock_prisma,
|
||||
patch(
|
||||
"backend.api.features.library.db.add_generated_agent_image",
|
||||
new=AsyncMock(),
|
||||
),
|
||||
patch(
|
||||
"backend.api.features.library.model.LibraryAgent.from_db",
|
||||
return_value=MagicMock(),
|
||||
),
|
||||
):
|
||||
mock_prisma.return_value.upsert = AsyncMock(return_value=mock_upserted)
|
||||
|
||||
result = await db.create_library_agent(mock_graph, "user-1")
|
||||
|
||||
assert len(result) == 1
|
||||
upsert_call = mock_prisma.return_value.upsert.call_args
|
||||
assert upsert_call is not None
|
||||
# Verify the upsert where clause uses the composite unique key
|
||||
where = upsert_call.kwargs["where"]
|
||||
assert "userId_agentGraphId_agentGraphVersion" in where
|
||||
# Verify the upsert data has both create and update branches
|
||||
data = upsert_call.kwargs["data"]
|
||||
assert "create" in data
|
||||
assert "update" in data
|
||||
# Verify update branch restores soft-deleted/archived agents
|
||||
assert data["update"]["isDeleted"] is False
|
||||
assert data["update"]["isArchived"] is False
|
||||
|
||||
@@ -12,7 +12,6 @@ Tests cover:
|
||||
5. Complete OAuth flow end-to-end
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
import base64
|
||||
import hashlib
|
||||
import secrets
|
||||
@@ -59,27 +58,14 @@ async def test_user(server, test_user_id: str):
|
||||
|
||||
yield test_user_id
|
||||
|
||||
# Cleanup - delete in correct order due to foreign key constraints.
|
||||
# Wrap in try/except because the event loop or Prisma engine may already
|
||||
# be closed during session teardown on Python 3.12+.
|
||||
try:
|
||||
await asyncio.gather(
|
||||
PrismaOAuthAccessToken.prisma().delete_many(where={"userId": test_user_id}),
|
||||
PrismaOAuthRefreshToken.prisma().delete_many(
|
||||
where={"userId": test_user_id}
|
||||
),
|
||||
PrismaOAuthAuthorizationCode.prisma().delete_many(
|
||||
where={"userId": test_user_id}
|
||||
),
|
||||
)
|
||||
await asyncio.gather(
|
||||
PrismaOAuthApplication.prisma().delete_many(
|
||||
where={"ownerId": test_user_id}
|
||||
),
|
||||
PrismaUser.prisma().delete(where={"id": test_user_id}),
|
||||
)
|
||||
except RuntimeError:
|
||||
pass
|
||||
# Cleanup - delete in correct order due to foreign key constraints
|
||||
await PrismaOAuthAccessToken.prisma().delete_many(where={"userId": test_user_id})
|
||||
await PrismaOAuthRefreshToken.prisma().delete_many(where={"userId": test_user_id})
|
||||
await PrismaOAuthAuthorizationCode.prisma().delete_many(
|
||||
where={"userId": test_user_id}
|
||||
)
|
||||
await PrismaOAuthApplication.prisma().delete_many(where={"ownerId": test_user_id})
|
||||
await PrismaUser.prisma().delete(where={"id": test_user_id})
|
||||
|
||||
|
||||
@pytest_asyncio.fixture
|
||||
|
||||
@@ -189,7 +189,6 @@ async def test_create_store_submission(mocker):
|
||||
notifyOnAgentApproved=True,
|
||||
notifyOnAgentRejected=True,
|
||||
timezone="Europe/Delft",
|
||||
subscriptionTier=prisma.enums.SubscriptionTier.FREE, # type: ignore[reportCallIssue,reportAttributeAccessIssue]
|
||||
)
|
||||
mock_agent = prisma.models.AgentGraph(
|
||||
id="agent-id",
|
||||
|
||||
@@ -18,8 +18,6 @@ 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
|
||||
import backend.api.features.builder.routes
|
||||
@@ -119,11 +117,6 @@ async def lifespan_context(app: fastapi.FastAPI):
|
||||
|
||||
AutoRegistry.patch_integrations()
|
||||
|
||||
# Register managed credential providers (e.g. AgentMail)
|
||||
from backend.integrations.managed_providers import register_all
|
||||
|
||||
register_all()
|
||||
|
||||
await backend.data.block.initialize_blocks()
|
||||
|
||||
await backend.data.user.migrate_and_encrypt_user_integrations()
|
||||
@@ -325,16 +318,6 @@ app.include_router(
|
||||
tags=["v2", "admin"],
|
||||
prefix="/api/executions",
|
||||
)
|
||||
app.include_router(
|
||||
backend.api.features.admin.rate_limit_admin_routes.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"],
|
||||
|
||||
@@ -698,30 +698,13 @@ class Block(ABC, Generic[BlockSchemaInputType, BlockSchemaOutputType]):
|
||||
if should_pause:
|
||||
return
|
||||
|
||||
# Validate the input data (original or reviewer-modified) once.
|
||||
# In dry-run mode, credential fields may contain sentinel None values
|
||||
# that would fail JSON schema required checks. We still validate the
|
||||
# non-credential fields so blocks that execute for real during dry-run
|
||||
# (e.g. AgentExecutorBlock) get proper input validation.
|
||||
is_dry_run = getattr(kwargs.get("execution_context"), "dry_run", False)
|
||||
if is_dry_run:
|
||||
cred_field_names = set(self.input_schema.get_credentials_fields().keys())
|
||||
non_cred_data = {
|
||||
k: v for k, v in input_data.items() if k not in cred_field_names
|
||||
}
|
||||
if error := self.input_schema.validate_data(non_cred_data):
|
||||
raise BlockInputError(
|
||||
message=f"Unable to execute block with invalid input data: {error}",
|
||||
block_name=self.name,
|
||||
block_id=self.id,
|
||||
)
|
||||
else:
|
||||
if error := self.input_schema.validate_data(input_data):
|
||||
raise BlockInputError(
|
||||
message=f"Unable to execute block with invalid input data: {error}",
|
||||
block_name=self.name,
|
||||
block_id=self.id,
|
||||
)
|
||||
# Validate the input data (original or reviewer-modified) once
|
||||
if error := self.input_schema.validate_data(input_data):
|
||||
raise BlockInputError(
|
||||
message=f"Unable to execute block with invalid input data: {error}",
|
||||
block_name=self.name,
|
||||
block_id=self.id,
|
||||
)
|
||||
|
||||
# Use the validated input data
|
||||
async for output_name, output_data in self.run(
|
||||
|
||||
@@ -49,17 +49,11 @@ class AgentExecutorBlock(Block):
|
||||
@classmethod
|
||||
def get_missing_input(cls, data: BlockInput) -> set[str]:
|
||||
required_fields = cls.get_input_schema(data).get("required", [])
|
||||
# Check against the nested `inputs` dict, not the top-level node
|
||||
# data — required fields like "topic" live inside data["inputs"],
|
||||
# not at data["topic"].
|
||||
provided = data.get("inputs", {})
|
||||
return set(required_fields) - set(provided)
|
||||
return set(required_fields) - set(data)
|
||||
|
||||
@classmethod
|
||||
def get_mismatch_error(cls, data: BlockInput) -> str | None:
|
||||
return validate_with_jsonschema(
|
||||
cls.get_input_schema(data), data.get("inputs", {})
|
||||
)
|
||||
return validate_with_jsonschema(cls.get_input_schema(data), data)
|
||||
|
||||
class Output(BlockSchema):
|
||||
# Use BlockSchema to avoid automatic error field that could clash with graph outputs
|
||||
@@ -94,7 +88,6 @@ class AgentExecutorBlock(Block):
|
||||
execution_context=execution_context.model_copy(
|
||||
update={"parent_execution_id": graph_exec_id},
|
||||
),
|
||||
dry_run=execution_context.dry_run,
|
||||
)
|
||||
|
||||
logger = execution_utils.LogMetadata(
|
||||
@@ -156,19 +149,14 @@ class AgentExecutorBlock(Block):
|
||||
ExecutionStatus.TERMINATED,
|
||||
ExecutionStatus.FAILED,
|
||||
]:
|
||||
logger.info(
|
||||
f"Execution {log_id} skipping event {event.event_type} status={event.status} "
|
||||
f"node={getattr(event, 'node_exec_id', '?')}"
|
||||
logger.debug(
|
||||
f"Execution {log_id} received event {event.event_type} with status {event.status}"
|
||||
)
|
||||
continue
|
||||
|
||||
if event.event_type == ExecutionEventType.GRAPH_EXEC_UPDATE:
|
||||
# If the graph execution is COMPLETED, TERMINATED, or FAILED,
|
||||
# we can stop listening for further events.
|
||||
logger.info(
|
||||
f"Execution {log_id} graph completed with status {event.status}, "
|
||||
f"yielded {len(yielded_node_exec_ids)} outputs"
|
||||
)
|
||||
self.merge_stats(
|
||||
NodeExecutionStats(
|
||||
extra_cost=event.stats.cost if event.stats else 0,
|
||||
|
||||
@@ -1,4 +1,3 @@
|
||||
import re
|
||||
from typing import Any
|
||||
|
||||
from backend.blocks._base import (
|
||||
@@ -20,33 +19,6 @@ from backend.blocks.llm import (
|
||||
)
|
||||
from backend.data.model import APIKeyCredentials, NodeExecutionStats, SchemaField
|
||||
|
||||
# Minimum max_output_tokens accepted by OpenAI-compatible APIs.
|
||||
# A true/false answer fits comfortably within this budget.
|
||||
MIN_LLM_OUTPUT_TOKENS = 16
|
||||
|
||||
|
||||
def _parse_boolean_response(response_text: str) -> tuple[bool, str | None]:
|
||||
"""Parse an LLM response into a boolean result.
|
||||
|
||||
Returns a ``(result, error)`` tuple. *error* is ``None`` when the
|
||||
response is unambiguous; otherwise it contains a diagnostic message
|
||||
and *result* defaults to ``False``.
|
||||
"""
|
||||
text = response_text.strip().lower()
|
||||
if text == "true":
|
||||
return True, None
|
||||
if text == "false":
|
||||
return False, None
|
||||
|
||||
# Fuzzy match – use word boundaries to avoid false positives like "untrue".
|
||||
tokens = set(re.findall(r"\b(true|false|yes|no|1|0)\b", text))
|
||||
if tokens == {"true"} or tokens == {"yes"} or tokens == {"1"}:
|
||||
return True, None
|
||||
if tokens == {"false"} or tokens == {"no"} or tokens == {"0"}:
|
||||
return False, None
|
||||
|
||||
return False, f"Unclear AI response: '{response_text}'"
|
||||
|
||||
|
||||
class AIConditionBlock(AIBlockBase):
|
||||
"""
|
||||
@@ -190,26 +162,54 @@ class AIConditionBlock(AIBlockBase):
|
||||
]
|
||||
|
||||
# Call the LLM
|
||||
response = await self.llm_call(
|
||||
credentials=credentials,
|
||||
llm_model=input_data.model,
|
||||
prompt=prompt,
|
||||
max_tokens=MIN_LLM_OUTPUT_TOKENS,
|
||||
)
|
||||
|
||||
# Extract the boolean result from the response
|
||||
result, error = _parse_boolean_response(response.response)
|
||||
if error:
|
||||
yield "error", error
|
||||
|
||||
# Update internal stats
|
||||
self.merge_stats(
|
||||
NodeExecutionStats(
|
||||
input_token_count=response.prompt_tokens,
|
||||
output_token_count=response.completion_tokens,
|
||||
try:
|
||||
response = await self.llm_call(
|
||||
credentials=credentials,
|
||||
llm_model=input_data.model,
|
||||
prompt=prompt,
|
||||
max_tokens=10, # We only expect a true/false response
|
||||
)
|
||||
)
|
||||
self.prompt = response.prompt
|
||||
|
||||
# Extract the boolean result from the response
|
||||
response_text = response.response.strip().lower()
|
||||
if response_text == "true":
|
||||
result = True
|
||||
elif response_text == "false":
|
||||
result = False
|
||||
else:
|
||||
# If the response is not clear, try to interpret it using word boundaries
|
||||
import re
|
||||
|
||||
# Use word boundaries to avoid false positives like 'untrue' or '10'
|
||||
tokens = set(re.findall(r"\b(true|false|yes|no|1|0)\b", response_text))
|
||||
|
||||
if tokens == {"true"} or tokens == {"yes"} or tokens == {"1"}:
|
||||
result = True
|
||||
elif tokens == {"false"} or tokens == {"no"} or tokens == {"0"}:
|
||||
result = False
|
||||
else:
|
||||
# Unclear or conflicting response - default to False and yield error
|
||||
result = False
|
||||
yield "error", f"Unclear AI response: '{response.response}'"
|
||||
|
||||
# Update internal stats
|
||||
self.merge_stats(
|
||||
NodeExecutionStats(
|
||||
input_token_count=response.prompt_tokens,
|
||||
output_token_count=response.completion_tokens,
|
||||
)
|
||||
)
|
||||
self.prompt = response.prompt
|
||||
|
||||
except Exception as e:
|
||||
# In case of any error, default to False to be safe
|
||||
result = False
|
||||
# Log the error but don't fail the block execution
|
||||
import logging
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
logger.error(f"AI condition evaluation failed: {str(e)}")
|
||||
yield "error", f"AI evaluation failed: {str(e)}"
|
||||
|
||||
# Yield results
|
||||
yield "result", result
|
||||
|
||||
@@ -1,147 +0,0 @@
|
||||
"""Tests for AIConditionBlock – regression coverage for max_tokens and error propagation."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from typing import cast
|
||||
|
||||
import pytest
|
||||
|
||||
from backend.blocks.ai_condition import (
|
||||
MIN_LLM_OUTPUT_TOKENS,
|
||||
AIConditionBlock,
|
||||
_parse_boolean_response,
|
||||
)
|
||||
from backend.blocks.llm import (
|
||||
DEFAULT_LLM_MODEL,
|
||||
TEST_CREDENTIALS,
|
||||
TEST_CREDENTIALS_INPUT,
|
||||
AICredentials,
|
||||
LLMResponse,
|
||||
)
|
||||
|
||||
_TEST_AI_CREDENTIALS = cast(AICredentials, TEST_CREDENTIALS_INPUT)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Helper to collect all yields from the async generator
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
async def _collect_outputs(block: AIConditionBlock, input_data, credentials):
|
||||
outputs: dict[str, object] = {}
|
||||
async for name, value in block.run(input_data, credentials=credentials):
|
||||
outputs[name] = value
|
||||
return outputs
|
||||
|
||||
|
||||
def _make_input(**overrides) -> AIConditionBlock.Input:
|
||||
defaults: dict = {
|
||||
"input_value": "hello@example.com",
|
||||
"condition": "the input is an email address",
|
||||
"yes_value": "yes!",
|
||||
"no_value": "no!",
|
||||
"model": DEFAULT_LLM_MODEL,
|
||||
"credentials": TEST_CREDENTIALS_INPUT,
|
||||
}
|
||||
defaults.update(overrides)
|
||||
return AIConditionBlock.Input(**defaults)
|
||||
|
||||
|
||||
def _mock_llm_response(response_text: str) -> LLMResponse:
|
||||
return LLMResponse(
|
||||
raw_response="",
|
||||
prompt=[],
|
||||
response=response_text,
|
||||
tool_calls=None,
|
||||
prompt_tokens=10,
|
||||
completion_tokens=5,
|
||||
reasoning=None,
|
||||
)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# _parse_boolean_response unit tests
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestParseBooleanResponse:
|
||||
def test_true_exact(self):
|
||||
assert _parse_boolean_response("true") == (True, None)
|
||||
|
||||
def test_false_exact(self):
|
||||
assert _parse_boolean_response("false") == (False, None)
|
||||
|
||||
def test_true_with_whitespace(self):
|
||||
assert _parse_boolean_response(" True ") == (True, None)
|
||||
|
||||
def test_yes_fuzzy(self):
|
||||
assert _parse_boolean_response("Yes") == (True, None)
|
||||
|
||||
def test_no_fuzzy(self):
|
||||
assert _parse_boolean_response("no") == (False, None)
|
||||
|
||||
def test_one_fuzzy(self):
|
||||
assert _parse_boolean_response("1") == (True, None)
|
||||
|
||||
def test_zero_fuzzy(self):
|
||||
assert _parse_boolean_response("0") == (False, None)
|
||||
|
||||
def test_unclear_response(self):
|
||||
result, error = _parse_boolean_response("I'm not sure")
|
||||
assert result is False
|
||||
assert error is not None
|
||||
assert "Unclear" in error
|
||||
|
||||
def test_conflicting_tokens(self):
|
||||
result, error = _parse_boolean_response("true and false")
|
||||
assert result is False
|
||||
assert error is not None
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Regression: max_tokens is set to MIN_LLM_OUTPUT_TOKENS
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestMaxTokensRegression:
|
||||
@pytest.mark.asyncio
|
||||
async def test_llm_call_receives_min_output_tokens(self):
|
||||
"""max_tokens must be MIN_LLM_OUTPUT_TOKENS (16) – the previous value
|
||||
of 1 was too low and caused OpenAI to reject the request."""
|
||||
block = AIConditionBlock()
|
||||
captured_kwargs: dict = {}
|
||||
|
||||
async def spy_llm_call(**kwargs):
|
||||
captured_kwargs.update(kwargs)
|
||||
return _mock_llm_response("true")
|
||||
|
||||
block.llm_call = spy_llm_call # type: ignore[assignment]
|
||||
|
||||
input_data = _make_input()
|
||||
await _collect_outputs(block, input_data, credentials=TEST_CREDENTIALS)
|
||||
|
||||
assert captured_kwargs["max_tokens"] == MIN_LLM_OUTPUT_TOKENS
|
||||
assert captured_kwargs["max_tokens"] == 16
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Regression: exceptions from llm_call must propagate
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestExceptionPropagation:
|
||||
@pytest.mark.asyncio
|
||||
async def test_llm_call_exception_propagates(self):
|
||||
"""If llm_call raises, the exception must NOT be swallowed.
|
||||
Previously the block caught all exceptions and silently returned
|
||||
result=False."""
|
||||
block = AIConditionBlock()
|
||||
|
||||
async def boom(**kwargs):
|
||||
raise RuntimeError("LLM provider error")
|
||||
|
||||
block.llm_call = boom # type: ignore[assignment]
|
||||
|
||||
input_data = _make_input()
|
||||
with pytest.raises(RuntimeError, match="LLM provider error"):
|
||||
await _collect_outputs(block, input_data, credentials=TEST_CREDENTIALS)
|
||||
@@ -18,7 +18,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -359,7 +358,6 @@ class AIShortformVideoCreatorBlock(Block):
|
||||
execution_context=execution_context,
|
||||
return_format="for_block_output",
|
||||
)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
yield "video_url", stored_url
|
||||
|
||||
|
||||
@@ -567,7 +565,6 @@ class AIAdMakerVideoCreatorBlock(Block):
|
||||
execution_context=execution_context,
|
||||
return_format="for_block_output",
|
||||
)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
yield "video_url", stored_url
|
||||
|
||||
|
||||
@@ -763,5 +760,4 @@ class AIScreenshotToVideoAdBlock(Block):
|
||||
execution_context=execution_context,
|
||||
return_format="for_block_output",
|
||||
)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
yield "video_url", stored_url
|
||||
|
||||
@@ -17,7 +17,7 @@ from backend.blocks.apollo.models import (
|
||||
PrimaryPhone,
|
||||
SearchOrganizationsRequest,
|
||||
)
|
||||
from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
|
||||
from backend.data.model import CredentialsField, SchemaField
|
||||
|
||||
|
||||
class SearchOrganizationsBlock(Block):
|
||||
@@ -218,7 +218,6 @@ 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
|
||||
|
||||
@@ -21,7 +21,7 @@ from backend.blocks.apollo.models import (
|
||||
SearchPeopleRequest,
|
||||
SenorityLevels,
|
||||
)
|
||||
from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
|
||||
from backend.data.model import CredentialsField, SchemaField
|
||||
|
||||
|
||||
class SearchPeopleBlock(Block):
|
||||
@@ -366,5 +366,4 @@ class SearchPeopleBlock(Block):
|
||||
*(enrich_or_fallback(person) for person in people)
|
||||
)
|
||||
|
||||
self.merge_stats(NodeExecutionStats(output_size=len(people)))
|
||||
yield "people", people
|
||||
|
||||
@@ -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, NodeExecutionStats, SchemaField
|
||||
from backend.data.model import CredentialsField, SchemaField
|
||||
|
||||
|
||||
class GetPersonDetailBlock(Block):
|
||||
@@ -141,5 +141,4 @@ 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)
|
||||
|
||||
@@ -146,21 +146,6 @@ class AutoPilotBlock(Block):
|
||||
advanced=True,
|
||||
)
|
||||
|
||||
dry_run: bool = SchemaField(
|
||||
description=(
|
||||
"When enabled, run_block and run_agent tool calls in this "
|
||||
"autopilot session are forced to use dry-run simulation mode. "
|
||||
"No real API calls, side effects, or credits are consumed "
|
||||
"by those tools. Useful for testing agent wiring and "
|
||||
"previewing outputs. "
|
||||
"Only applies when creating a new session (session_id is empty). "
|
||||
"When reusing an existing session_id, the session's original "
|
||||
"dry_run setting is preserved."
|
||||
),
|
||||
default=False,
|
||||
advanced=True,
|
||||
)
|
||||
|
||||
# timeout_seconds removed: the SDK manages its own heartbeat-based
|
||||
# timeouts internally; wrapping with asyncio.timeout corrupts the
|
||||
# SDK's internal stream (see service.py CRITICAL comment).
|
||||
@@ -247,11 +232,11 @@ class AutoPilotBlock(Block):
|
||||
},
|
||||
)
|
||||
|
||||
async def create_session(self, user_id: str, *, dry_run: bool) -> str:
|
||||
async def create_session(self, user_id: str) -> str:
|
||||
"""Create a new chat session and return its ID (mockable for tests)."""
|
||||
from backend.copilot.model import create_chat_session # avoid circular import
|
||||
|
||||
session = await create_chat_session(user_id, dry_run=dry_run)
|
||||
session = await create_chat_session(user_id)
|
||||
return session.session_id
|
||||
|
||||
async def execute_copilot(
|
||||
@@ -382,9 +367,7 @@ class AutoPilotBlock(Block):
|
||||
# even if the downstream stream fails (avoids orphaned sessions).
|
||||
sid = input_data.session_id
|
||||
if not sid:
|
||||
sid = await self.create_session(
|
||||
execution_context.user_id, dry_run=input_data.dry_run
|
||||
)
|
||||
sid = await self.create_session(execution_context.user_id)
|
||||
|
||||
# NOTE: No asyncio.timeout() here — the SDK manages its own
|
||||
# heartbeat-based timeouts internally. Wrapping with asyncio.timeout
|
||||
|
||||
@@ -17,7 +17,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -343,7 +342,6 @@ 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
|
||||
@@ -469,7 +467,6 @@ 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:
|
||||
@@ -580,7 +577,6 @@ 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
|
||||
|
||||
@@ -73,7 +73,7 @@ class ReadDiscordMessagesBlock(Block):
|
||||
id="df06086a-d5ac-4abb-9996-2ad0acb2eff7",
|
||||
input_schema=ReadDiscordMessagesBlock.Input, # Assign input schema
|
||||
output_schema=ReadDiscordMessagesBlock.Output, # Assign output schema
|
||||
description="Reads new messages from a Discord channel using a bot token and triggers when a new message is posted",
|
||||
description="Reads messages from a Discord channel using a bot token.",
|
||||
categories={BlockCategory.SOCIAL},
|
||||
test_input={
|
||||
"continuous_read": False,
|
||||
|
||||
@@ -15,12 +15,7 @@ from backend.blocks._base import (
|
||||
BlockSchemaInput,
|
||||
BlockSchemaOutput,
|
||||
)
|
||||
from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.data.model import APIKeyCredentials, CredentialsField, SchemaField
|
||||
from backend.util.type import MediaFileType
|
||||
|
||||
from ._api import (
|
||||
@@ -200,7 +195,6 @@ 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)}")
|
||||
@@ -347,7 +341,6 @@ 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)}")
|
||||
@@ -450,7 +443,6 @@ 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)}")
|
||||
@@ -531,7 +523,6 @@ 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)}")
|
||||
|
||||
@@ -4,7 +4,6 @@ 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,
|
||||
@@ -224,6 +223,3 @@ class ExaContentsBlock(Block):
|
||||
|
||||
if response.cost_dollars:
|
||||
yield "cost_dollars", response.cost_dollars
|
||||
self.merge_stats(
|
||||
NodeExecutionStats(provider_cost=response.cost_dollars.total)
|
||||
)
|
||||
|
||||
@@ -4,7 +4,6 @@ from typing import Optional
|
||||
|
||||
from exa_py import AsyncExa
|
||||
|
||||
from backend.data.model import NodeExecutionStats
|
||||
from backend.sdk import (
|
||||
APIKeyCredentials,
|
||||
Block,
|
||||
@@ -207,6 +206,3 @@ class ExaSearchBlock(Block):
|
||||
|
||||
if response.cost_dollars:
|
||||
yield "cost_dollars", response.cost_dollars
|
||||
self.merge_stats(
|
||||
NodeExecutionStats(provider_cost=response.cost_dollars.total)
|
||||
)
|
||||
|
||||
@@ -18,7 +18,7 @@ from backend.blocks.fal._auth import (
|
||||
FalCredentialsInput,
|
||||
)
|
||||
from backend.data.execution import ExecutionContext
|
||||
from backend.data.model import NodeExecutionStats, SchemaField
|
||||
from backend.data.model import SchemaField
|
||||
from backend.util.file import store_media_file
|
||||
from backend.util.request import ClientResponseError, Requests
|
||||
from backend.util.type import MediaFileType
|
||||
@@ -230,7 +230,6 @@ 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)
|
||||
|
||||
@@ -1,6 +1,5 @@
|
||||
import asyncio
|
||||
import base64
|
||||
import re
|
||||
from abc import ABC
|
||||
from email import encoders
|
||||
from email.mime.base import MIMEBase
|
||||
@@ -9,7 +8,7 @@ from email.mime.text import MIMEText
|
||||
from email.policy import SMTP
|
||||
from email.utils import getaddresses, parseaddr
|
||||
from pathlib import Path
|
||||
from typing import List, Literal, Optional, Protocol, runtime_checkable
|
||||
from typing import List, Literal, Optional
|
||||
|
||||
from google.oauth2.credentials import Credentials
|
||||
from googleapiclient.discovery import build
|
||||
@@ -43,52 +42,8 @@ NO_WRAP_POLICY = SMTP.clone(max_line_length=0)
|
||||
|
||||
|
||||
def serialize_email_recipients(recipients: list[str]) -> str:
|
||||
"""Serialize recipients list to comma-separated string.
|
||||
|
||||
Strips leading/trailing whitespace from each address to keep MIME
|
||||
headers clean (mirrors the strip done in ``validate_email_recipients``).
|
||||
"""
|
||||
return ", ".join(addr.strip() for addr in recipients)
|
||||
|
||||
|
||||
# RFC 5322 simplified pattern: local@domain where domain has at least one dot
|
||||
_EMAIL_RE = re.compile(r"^[^@\s]+@[^@\s]+\.[^@\s]+$")
|
||||
|
||||
|
||||
def validate_email_recipients(recipients: list[str], field_name: str = "to") -> None:
|
||||
"""Validate that all recipients are plausible email addresses.
|
||||
|
||||
Raises ``ValueError`` with a user-friendly message listing every
|
||||
invalid entry so the caller (or LLM) can correct them in one pass.
|
||||
"""
|
||||
invalid = [addr for addr in recipients if not _EMAIL_RE.match(addr.strip())]
|
||||
if invalid:
|
||||
formatted = ", ".join(f"'{a}'" for a in invalid)
|
||||
raise ValueError(
|
||||
f"Invalid email address(es) in '{field_name}': {formatted}. "
|
||||
f"Each entry must be a valid email address (e.g. user@example.com)."
|
||||
)
|
||||
|
||||
|
||||
@runtime_checkable
|
||||
class HasRecipients(Protocol):
|
||||
to: list[str]
|
||||
cc: list[str]
|
||||
bcc: list[str]
|
||||
|
||||
|
||||
def validate_all_recipients(input_data: HasRecipients) -> None:
|
||||
"""Validate to/cc/bcc recipient fields on an input namespace.
|
||||
|
||||
Calls ``validate_email_recipients`` for ``to`` (required) and
|
||||
``cc``/``bcc`` (when non-empty), raising ``ValueError`` on the
|
||||
first field that contains an invalid address.
|
||||
"""
|
||||
validate_email_recipients(input_data.to, "to")
|
||||
if input_data.cc:
|
||||
validate_email_recipients(input_data.cc, "cc")
|
||||
if input_data.bcc:
|
||||
validate_email_recipients(input_data.bcc, "bcc")
|
||||
"""Serialize recipients list to comma-separated string."""
|
||||
return ", ".join(recipients)
|
||||
|
||||
|
||||
def _make_mime_text(
|
||||
@@ -145,16 +100,14 @@ async def create_mime_message(
|
||||
) -> str:
|
||||
"""Create a MIME message with attachments and return base64-encoded raw message."""
|
||||
|
||||
validate_all_recipients(input_data)
|
||||
|
||||
message = MIMEMultipart()
|
||||
message["to"] = serialize_email_recipients(input_data.to)
|
||||
message["subject"] = input_data.subject
|
||||
|
||||
if input_data.cc:
|
||||
message["cc"] = serialize_email_recipients(input_data.cc)
|
||||
message["cc"] = ", ".join(input_data.cc)
|
||||
if input_data.bcc:
|
||||
message["bcc"] = serialize_email_recipients(input_data.bcc)
|
||||
message["bcc"] = ", ".join(input_data.bcc)
|
||||
|
||||
# Use the new helper function with content_type if available
|
||||
content_type = getattr(input_data, "content_type", None)
|
||||
@@ -1214,15 +1167,13 @@ async def _build_reply_message(
|
||||
references.append(headers["message-id"])
|
||||
|
||||
# Create MIME message
|
||||
validate_all_recipients(input_data)
|
||||
|
||||
msg = MIMEMultipart()
|
||||
if input_data.to:
|
||||
msg["To"] = serialize_email_recipients(input_data.to)
|
||||
msg["To"] = ", ".join(input_data.to)
|
||||
if input_data.cc:
|
||||
msg["Cc"] = serialize_email_recipients(input_data.cc)
|
||||
msg["Cc"] = ", ".join(input_data.cc)
|
||||
if input_data.bcc:
|
||||
msg["Bcc"] = serialize_email_recipients(input_data.bcc)
|
||||
msg["Bcc"] = ", ".join(input_data.bcc)
|
||||
msg["Subject"] = subject
|
||||
if headers.get("message-id"):
|
||||
msg["In-Reply-To"] = headers["message-id"]
|
||||
@@ -1734,16 +1685,13 @@ To: {original_to}
|
||||
else:
|
||||
body = f"{forward_header}\n\n{original_body}"
|
||||
|
||||
# Validate all recipient lists before building the MIME message
|
||||
validate_all_recipients(input_data)
|
||||
|
||||
# Create MIME message
|
||||
msg = MIMEMultipart()
|
||||
msg["To"] = serialize_email_recipients(input_data.to)
|
||||
msg["To"] = ", ".join(input_data.to)
|
||||
if input_data.cc:
|
||||
msg["Cc"] = serialize_email_recipients(input_data.cc)
|
||||
msg["Cc"] = ", ".join(input_data.cc)
|
||||
if input_data.bcc:
|
||||
msg["Bcc"] = serialize_email_recipients(input_data.bcc)
|
||||
msg["Bcc"] = ", ".join(input_data.bcc)
|
||||
msg["Subject"] = subject
|
||||
|
||||
# Add body with proper content type
|
||||
|
||||
@@ -14,7 +14,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -118,7 +117,6 @@ 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
|
||||
|
||||
|
||||
@@ -14,7 +14,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -228,7 +227,6 @@ class IdeogramModelBlock(Block):
|
||||
image_url=result,
|
||||
)
|
||||
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
yield "result", result
|
||||
|
||||
async def run_model(
|
||||
|
||||
@@ -2,8 +2,6 @@ import copy
|
||||
from datetime import date, time
|
||||
from typing import Any, Optional
|
||||
|
||||
from pydantic import AliasChoices, Field
|
||||
|
||||
from backend.blocks._base import (
|
||||
Block,
|
||||
BlockCategory,
|
||||
@@ -30,9 +28,9 @@ class AgentInputBlock(Block):
|
||||
"""
|
||||
This block is used to provide input to the graph.
|
||||
|
||||
It takes in a value, name, and description.
|
||||
It takes in a value, name, description, default values list and bool to limit selection to default values.
|
||||
|
||||
It outputs the value passed as input.
|
||||
It Outputs the value passed as input.
|
||||
"""
|
||||
|
||||
class Input(BlockSchemaInput):
|
||||
@@ -49,6 +47,12 @@ class AgentInputBlock(Block):
|
||||
default=None,
|
||||
advanced=True,
|
||||
)
|
||||
placeholder_values: list = SchemaField(
|
||||
description="The placeholder values to be passed as input.",
|
||||
default_factory=list,
|
||||
advanced=True,
|
||||
hidden=True,
|
||||
)
|
||||
advanced: bool = SchemaField(
|
||||
description="Whether to show the input in the advanced section, if the field is not required.",
|
||||
default=False,
|
||||
@@ -61,7 +65,10 @@ class AgentInputBlock(Block):
|
||||
)
|
||||
|
||||
def generate_schema(self):
|
||||
return copy.deepcopy(self.get_field_schema("value"))
|
||||
schema = copy.deepcopy(self.get_field_schema("value"))
|
||||
if possible_values := self.placeholder_values:
|
||||
schema["enum"] = possible_values
|
||||
return schema
|
||||
|
||||
class Output(BlockSchema):
|
||||
# Use BlockSchema to avoid automatic error field for interface definition
|
||||
@@ -79,16 +86,18 @@ class AgentInputBlock(Block):
|
||||
"value": "Hello, World!",
|
||||
"name": "input_1",
|
||||
"description": "Example test input.",
|
||||
"placeholder_values": [],
|
||||
},
|
||||
{
|
||||
"value": 42,
|
||||
"value": "Hello, World!",
|
||||
"name": "input_2",
|
||||
"description": "Example numeric input.",
|
||||
"description": "Example test input with placeholders.",
|
||||
"placeholder_values": ["Hello, World!"],
|
||||
},
|
||||
],
|
||||
"test_output": [
|
||||
("result", "Hello, World!"),
|
||||
("result", 42),
|
||||
("result", "Hello, World!"),
|
||||
],
|
||||
"categories": {BlockCategory.INPUT, BlockCategory.BASIC},
|
||||
"block_type": BlockType.INPUT,
|
||||
@@ -236,11 +245,13 @@ class AgentShortTextInputBlock(AgentInputBlock):
|
||||
"value": "Hello",
|
||||
"name": "short_text_1",
|
||||
"description": "Short text example 1",
|
||||
"placeholder_values": [],
|
||||
},
|
||||
{
|
||||
"value": "Quick test",
|
||||
"name": "short_text_2",
|
||||
"description": "Short text example 2",
|
||||
"placeholder_values": ["Quick test", "Another option"],
|
||||
},
|
||||
],
|
||||
test_output=[
|
||||
@@ -274,11 +285,13 @@ class AgentLongTextInputBlock(AgentInputBlock):
|
||||
"value": "Lorem ipsum dolor sit amet...",
|
||||
"name": "long_text_1",
|
||||
"description": "Long text example 1",
|
||||
"placeholder_values": [],
|
||||
},
|
||||
{
|
||||
"value": "Another multiline text input.",
|
||||
"name": "long_text_2",
|
||||
"description": "Long text example 2",
|
||||
"placeholder_values": ["Another multiline text input."],
|
||||
},
|
||||
],
|
||||
test_output=[
|
||||
@@ -312,11 +325,13 @@ class AgentNumberInputBlock(AgentInputBlock):
|
||||
"value": 42,
|
||||
"name": "number_input_1",
|
||||
"description": "Number example 1",
|
||||
"placeholder_values": [],
|
||||
},
|
||||
{
|
||||
"value": 314,
|
||||
"name": "number_input_2",
|
||||
"description": "Number example 2",
|
||||
"placeholder_values": [314, 2718],
|
||||
},
|
||||
],
|
||||
test_output=[
|
||||
@@ -469,8 +484,7 @@ class AgentFileInputBlock(AgentInputBlock):
|
||||
|
||||
class AgentDropdownInputBlock(AgentInputBlock):
|
||||
"""
|
||||
A specialized text input block that presents a dropdown selector
|
||||
restricted to a fixed set of values.
|
||||
A specialized text input block that relies on placeholder_values to present a dropdown.
|
||||
"""
|
||||
|
||||
class Input(AgentInputBlock.Input):
|
||||
@@ -480,26 +494,13 @@ class AgentDropdownInputBlock(AgentInputBlock):
|
||||
advanced=False,
|
||||
title="Default Value",
|
||||
)
|
||||
# Use Field() directly (not SchemaField) to pass validation_alias,
|
||||
# which handles backward compat for legacy "placeholder_values" across
|
||||
# all construction paths (model_construct, __init__, model_validate).
|
||||
options: list = Field(
|
||||
placeholder_values: list = SchemaField(
|
||||
description="Possible values for the dropdown.",
|
||||
default_factory=list,
|
||||
advanced=False,
|
||||
title="Dropdown Options",
|
||||
description=(
|
||||
"If provided, renders the input as a dropdown selector "
|
||||
"restricted to these values. Leave empty for free-text input."
|
||||
),
|
||||
validation_alias=AliasChoices("options", "placeholder_values"),
|
||||
json_schema_extra={"advanced": False, "secret": False},
|
||||
)
|
||||
|
||||
def generate_schema(self):
|
||||
schema = super().generate_schema()
|
||||
if possible_values := self.options:
|
||||
schema["enum"] = possible_values
|
||||
return schema
|
||||
|
||||
class Output(AgentInputBlock.Output):
|
||||
result: str = SchemaField(description="Selected dropdown value.")
|
||||
|
||||
@@ -514,13 +515,13 @@ class AgentDropdownInputBlock(AgentInputBlock):
|
||||
{
|
||||
"value": "Option A",
|
||||
"name": "dropdown_1",
|
||||
"options": ["Option A", "Option B", "Option C"],
|
||||
"placeholder_values": ["Option A", "Option B", "Option C"],
|
||||
"description": "Dropdown example 1",
|
||||
},
|
||||
{
|
||||
"value": "Option C",
|
||||
"name": "dropdown_2",
|
||||
"options": ["Option A", "Option B", "Option C"],
|
||||
"placeholder_values": ["Option A", "Option B", "Option C"],
|
||||
"description": "Dropdown example 2",
|
||||
},
|
||||
],
|
||||
|
||||
@@ -10,7 +10,7 @@ from backend.blocks.jina._auth import (
|
||||
JinaCredentialsField,
|
||||
JinaCredentialsInput,
|
||||
)
|
||||
from backend.data.model import NodeExecutionStats, SchemaField
|
||||
from backend.data.model import SchemaField
|
||||
from backend.util.request import Requests
|
||||
|
||||
|
||||
@@ -45,13 +45,5 @@ class JinaEmbeddingBlock(Block):
|
||||
}
|
||||
data = {"input": input_data.texts, "model": input_data.model}
|
||||
response = await Requests().post(url, headers=headers, 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),
|
||||
)
|
||||
)
|
||||
embeddings = [e["embedding"] for e in response.json()["data"]]
|
||||
yield "embeddings", embeddings
|
||||
|
||||
@@ -104,18 +104,6 @@ class LlmModelMeta(EnumMeta):
|
||||
|
||||
|
||||
class LlmModel(str, Enum, metaclass=LlmModelMeta):
|
||||
|
||||
@classmethod
|
||||
def _missing_(cls, value: object) -> "LlmModel | None":
|
||||
"""Handle provider-prefixed model names like 'anthropic/claude-sonnet-4-6'."""
|
||||
if isinstance(value, str) and "/" in value:
|
||||
stripped = value.split("/", 1)[1]
|
||||
try:
|
||||
return cls(stripped)
|
||||
except ValueError:
|
||||
return None
|
||||
return None
|
||||
|
||||
# OpenAI models
|
||||
O3_MINI = "o3-mini"
|
||||
O3 = "o3-2025-04-16"
|
||||
@@ -687,7 +675,6 @@ 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(
|
||||
@@ -725,9 +712,6 @@ def convert_openai_tool_fmt_to_anthropic(
|
||||
def extract_openai_reasoning(response) -> str | None:
|
||||
"""Extract reasoning from OpenAI-compatible response if available."""
|
||||
"""Note: This will likely not working since the reasoning is not present in another Response API"""
|
||||
if not response.choices:
|
||||
logger.warning("LLM response has empty choices in extract_openai_reasoning")
|
||||
return None
|
||||
reasoning = None
|
||||
choice = response.choices[0]
|
||||
if hasattr(choice, "reasoning") and getattr(choice, "reasoning", None):
|
||||
@@ -743,9 +727,6 @@ def extract_openai_reasoning(response) -> str | None:
|
||||
|
||||
def extract_openai_tool_calls(response) -> list[ToolContentBlock] | None:
|
||||
"""Extract tool calls from OpenAI-compatible response."""
|
||||
if not response.choices:
|
||||
logger.warning("LLM response has empty choices in extract_openai_tool_calls")
|
||||
return None
|
||||
if response.choices[0].message.tool_calls:
|
||||
return [
|
||||
ToolContentBlock(
|
||||
@@ -979,8 +960,6 @@ async def llm_call(
|
||||
response_format=response_format, # type: ignore
|
||||
max_tokens=max_tokens,
|
||||
)
|
||||
if not response.choices:
|
||||
raise ValueError("Groq returned empty choices in response")
|
||||
return LLMResponse(
|
||||
raw_response=response.choices[0].message,
|
||||
prompt=prompt,
|
||||
@@ -1040,22 +1019,16 @@ async def llm_call(
|
||||
parallel_tool_calls=parallel_tool_calls_param,
|
||||
)
|
||||
|
||||
# If there's no response, raise an error
|
||||
if not response.choices:
|
||||
raise ValueError(f"OpenRouter returned empty choices: {response}")
|
||||
if response:
|
||||
raise ValueError(f"OpenRouter error: {response}")
|
||||
else:
|
||||
raise ValueError("No response from OpenRouter.")
|
||||
|
||||
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,
|
||||
@@ -1064,7 +1037,6 @@ 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
|
||||
@@ -1089,8 +1061,12 @@ async def llm_call(
|
||||
parallel_tool_calls=parallel_tool_calls_param,
|
||||
)
|
||||
|
||||
# If there's no response, raise an error
|
||||
if not response.choices:
|
||||
raise ValueError(f"Llama API returned empty choices: {response}")
|
||||
if response:
|
||||
raise ValueError(f"Llama API error: {response}")
|
||||
else:
|
||||
raise ValueError("No response from Llama API.")
|
||||
|
||||
tool_calls = extract_openai_tool_calls(response)
|
||||
reasoning = extract_openai_reasoning(response)
|
||||
@@ -1120,8 +1096,6 @@ async def llm_call(
|
||||
messages=prompt, # type: ignore
|
||||
max_tokens=max_tokens,
|
||||
)
|
||||
if not completion.choices:
|
||||
raise ValueError("AI/ML API returned empty choices in response")
|
||||
|
||||
return LLMResponse(
|
||||
raw_response=completion.choices[0].message,
|
||||
@@ -1158,9 +1132,6 @@ async def llm_call(
|
||||
parallel_tool_calls=parallel_tool_calls_param,
|
||||
)
|
||||
|
||||
if not response.choices:
|
||||
raise ValueError(f"v0 API returned empty choices: {response}")
|
||||
|
||||
tool_calls = extract_openai_tool_calls(response)
|
||||
reasoning = extract_openai_reasoning(response)
|
||||
|
||||
@@ -1389,13 +1360,12 @@ class AIStructuredResponseGeneratorBlock(AIBlockBase):
|
||||
max_tokens=input_data.max_tokens,
|
||||
)
|
||||
response_text = llm_response.response
|
||||
cost_stats = NodeExecutionStats(
|
||||
input_token_count=llm_response.prompt_tokens,
|
||||
output_token_count=llm_response.completion_tokens,
|
||||
self.merge_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:
|
||||
@@ -2029,19 +1999,6 @@ class AIConversationBlock(AIBlockBase):
|
||||
async def run(
|
||||
self, input_data: Input, *, credentials: APIKeyCredentials, **kwargs
|
||||
) -> BlockOutput:
|
||||
has_messages = any(
|
||||
isinstance(m, dict)
|
||||
and isinstance(m.get("content"), str)
|
||||
and bool(m["content"].strip())
|
||||
for m in (input_data.messages or [])
|
||||
)
|
||||
has_prompt = bool(input_data.prompt and input_data.prompt.strip())
|
||||
if not has_messages and not has_prompt:
|
||||
raise ValueError(
|
||||
"Cannot call LLM with no messages and no prompt. "
|
||||
"Provide at least one message or a non-empty prompt."
|
||||
)
|
||||
|
||||
response = await self.llm_call(
|
||||
AIStructuredResponseGeneratorBlock.Input(
|
||||
prompt=input_data.prompt,
|
||||
|
||||
@@ -89,12 +89,6 @@ class MCPToolBlock(Block):
|
||||
default={},
|
||||
hidden=True,
|
||||
)
|
||||
tool_description: str = SchemaField(
|
||||
description="Description of the selected MCP tool. "
|
||||
"Populated automatically when a tool is selected.",
|
||||
default="",
|
||||
hidden=True,
|
||||
)
|
||||
|
||||
tool_arguments: dict[str, Any] = SchemaField(
|
||||
description="Arguments to pass to the selected MCP tool. "
|
||||
|
||||
@@ -8,7 +8,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -154,7 +153,6 @@ class AddMemoryBlock(Block, Mem0Base):
|
||||
messages,
|
||||
**params,
|
||||
)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
|
||||
results = result.get("results", [])
|
||||
yield "results", results
|
||||
@@ -257,7 +255,6 @@ 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:
|
||||
@@ -343,7 +340,6 @@ class GetAllMemoriesBlock(Block, Mem0Base):
|
||||
filters=filters,
|
||||
version="v2",
|
||||
)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
|
||||
yield "memories", memories
|
||||
|
||||
@@ -438,7 +434,6 @@ 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)
|
||||
|
||||
@@ -10,7 +10,7 @@ from backend.blocks.nvidia._auth import (
|
||||
NvidiaCredentialsField,
|
||||
NvidiaCredentialsInput,
|
||||
)
|
||||
from backend.data.model import NodeExecutionStats, SchemaField
|
||||
from backend.data.model import SchemaField
|
||||
from backend.util.request import Requests
|
||||
from backend.util.type import MediaFileType
|
||||
|
||||
@@ -69,7 +69,6 @@ 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
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -17,12 +17,7 @@ from backend.blocks.replicate._auth import (
|
||||
ReplicateCredentialsInput,
|
||||
)
|
||||
from backend.blocks.replicate._helper import ReplicateOutputs, extract_result
|
||||
from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.data.model import APIKeyCredentials, CredentialsField, SchemaField
|
||||
from backend.util.exceptions import BlockExecutionError, BlockInputError
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
@@ -113,7 +108,6 @@ 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
|
||||
|
||||
@@ -16,7 +16,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -186,7 +185,6 @@ 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)
|
||||
|
||||
@@ -15,7 +15,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -147,7 +146,6 @@ 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"]
|
||||
|
||||
@@ -23,7 +23,7 @@ from backend.blocks.smartlead.models import (
|
||||
SaveSequencesResponse,
|
||||
Sequence,
|
||||
)
|
||||
from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
|
||||
from backend.data.model import CredentialsField, SchemaField
|
||||
|
||||
|
||||
class CreateCampaignBlock(Block):
|
||||
@@ -100,7 +100,6 @@ 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
|
||||
@@ -227,7 +226,6 @@ 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
|
||||
@@ -323,7 +321,6 @@ 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
|
||||
|
||||
@@ -1,304 +0,0 @@
|
||||
import asyncio
|
||||
from typing import Any, Literal
|
||||
|
||||
from pydantic import SecretStr
|
||||
from sqlalchemy.engine.url import URL
|
||||
from sqlalchemy.exc import DBAPIError, OperationalError, ProgrammingError
|
||||
|
||||
from backend.blocks._base import (
|
||||
Block,
|
||||
BlockCategory,
|
||||
BlockOutput,
|
||||
BlockSchemaInput,
|
||||
BlockSchemaOutput,
|
||||
)
|
||||
from backend.blocks.sql_query_helpers import (
|
||||
_DATABASE_TYPE_DEFAULT_PORT,
|
||||
_DATABASE_TYPE_TO_DRIVER,
|
||||
DatabaseType,
|
||||
_execute_query,
|
||||
_sanitize_error,
|
||||
_validate_query_is_read_only,
|
||||
_validate_single_statement,
|
||||
)
|
||||
from backend.data.model import (
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
SchemaField,
|
||||
UserPasswordCredentials,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
from backend.util.request import resolve_and_check_blocked
|
||||
|
||||
TEST_CREDENTIALS = UserPasswordCredentials(
|
||||
id="01234567-89ab-cdef-0123-456789abcdef",
|
||||
provider="database",
|
||||
username=SecretStr("test_user"),
|
||||
password=SecretStr("test_pass"),
|
||||
title="Mock Database credentials",
|
||||
)
|
||||
|
||||
TEST_CREDENTIALS_INPUT = {
|
||||
"provider": TEST_CREDENTIALS.provider,
|
||||
"id": TEST_CREDENTIALS.id,
|
||||
"type": TEST_CREDENTIALS.type,
|
||||
"title": TEST_CREDENTIALS.title,
|
||||
}
|
||||
|
||||
DatabaseCredentials = UserPasswordCredentials
|
||||
DatabaseCredentialsInput = CredentialsMetaInput[
|
||||
Literal[ProviderName.DATABASE],
|
||||
Literal["user_password"],
|
||||
]
|
||||
|
||||
|
||||
def DatabaseCredentialsField() -> DatabaseCredentialsInput:
|
||||
return CredentialsField(
|
||||
description="Database username and password",
|
||||
)
|
||||
|
||||
|
||||
class SQLQueryBlock(Block):
|
||||
class Input(BlockSchemaInput):
|
||||
database_type: DatabaseType = SchemaField(
|
||||
default=DatabaseType.POSTGRES,
|
||||
description="Database engine",
|
||||
advanced=False,
|
||||
)
|
||||
host: SecretStr = SchemaField(
|
||||
description="Database hostname or IP address",
|
||||
placeholder="db.example.com",
|
||||
secret=True,
|
||||
)
|
||||
port: int | None = SchemaField(
|
||||
default=None,
|
||||
description=(
|
||||
"Database port (leave empty for default: "
|
||||
"PostgreSQL: 5432, MySQL: 3306, MSSQL: 1433)"
|
||||
),
|
||||
ge=1,
|
||||
le=65535,
|
||||
)
|
||||
database: str = SchemaField(
|
||||
description="Name of the database to connect to",
|
||||
placeholder="my_database",
|
||||
)
|
||||
query: str = SchemaField(
|
||||
description="SQL query to execute",
|
||||
placeholder="SELECT * FROM analytics.daily_active_users LIMIT 10",
|
||||
)
|
||||
read_only: bool = SchemaField(
|
||||
default=True,
|
||||
description=(
|
||||
"When enabled (default), only SELECT queries are allowed "
|
||||
"and the database session is set to read-only mode. "
|
||||
"Disable to allow write operations (INSERT, UPDATE, DELETE, etc.)."
|
||||
),
|
||||
)
|
||||
timeout: int = SchemaField(
|
||||
default=30,
|
||||
description="Query timeout in seconds (max 120)",
|
||||
ge=1,
|
||||
le=120,
|
||||
)
|
||||
max_rows: int = SchemaField(
|
||||
default=1000,
|
||||
description="Maximum number of rows to return (max 10000)",
|
||||
ge=1,
|
||||
le=10000,
|
||||
)
|
||||
credentials: DatabaseCredentialsInput = DatabaseCredentialsField()
|
||||
|
||||
class Output(BlockSchemaOutput):
|
||||
results: list[dict[str, Any]] = SchemaField(
|
||||
description="Query results as a list of row dictionaries"
|
||||
)
|
||||
columns: list[str] = SchemaField(
|
||||
description="Column names from the query result"
|
||||
)
|
||||
row_count: int = SchemaField(description="Number of rows returned")
|
||||
affected_rows: int = SchemaField(
|
||||
description="Number of rows affected by a write query (INSERT/UPDATE/DELETE)"
|
||||
)
|
||||
error: str = SchemaField(description="Error message if the query failed")
|
||||
|
||||
def __init__(self):
|
||||
super().__init__(
|
||||
id="4dc35c0f-4fd8-465e-9616-5a216f1ba2bc",
|
||||
description=(
|
||||
"Execute a SQL query. Read-only by default for safety "
|
||||
"-- disable to allow write operations. "
|
||||
"Supports PostgreSQL, MySQL, and MSSQL via SQLAlchemy."
|
||||
),
|
||||
categories={BlockCategory.DATA},
|
||||
input_schema=SQLQueryBlock.Input,
|
||||
output_schema=SQLQueryBlock.Output,
|
||||
test_input={
|
||||
"query": "SELECT 1 AS test_col",
|
||||
"database_type": DatabaseType.POSTGRES,
|
||||
"host": "localhost",
|
||||
"database": "test_db",
|
||||
"timeout": 30,
|
||||
"max_rows": 1000,
|
||||
"credentials": TEST_CREDENTIALS_INPUT,
|
||||
},
|
||||
test_credentials=TEST_CREDENTIALS,
|
||||
test_output=[
|
||||
("results", [{"test_col": 1}]),
|
||||
("columns", ["test_col"]),
|
||||
("row_count", 1),
|
||||
],
|
||||
test_mock={
|
||||
"execute_query": lambda *_args, **_kwargs: (
|
||||
[{"test_col": 1}],
|
||||
["test_col"],
|
||||
-1,
|
||||
),
|
||||
"check_host_allowed": lambda *_args, **_kwargs: ["127.0.0.1"],
|
||||
},
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
async def check_host_allowed(host: str) -> list[str]:
|
||||
"""Validate that the given host is not a private/blocked address.
|
||||
|
||||
Returns the list of resolved IP addresses so the caller can pin the
|
||||
connection to the validated IP (preventing DNS rebinding / TOCTOU).
|
||||
Raises ValueError or OSError if the host is blocked.
|
||||
Extracted as a method so it can be mocked during block tests.
|
||||
"""
|
||||
return await resolve_and_check_blocked(host)
|
||||
|
||||
@staticmethod
|
||||
def execute_query(
|
||||
connection_url: URL | str,
|
||||
query: str,
|
||||
timeout: int,
|
||||
max_rows: int,
|
||||
read_only: bool = True,
|
||||
database_type: DatabaseType = DatabaseType.POSTGRES,
|
||||
) -> 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.
|
||||
"""
|
||||
return _execute_query(
|
||||
connection_url=connection_url,
|
||||
query=query,
|
||||
timeout=timeout,
|
||||
max_rows=max_rows,
|
||||
read_only=read_only,
|
||||
database_type=database_type,
|
||||
)
|
||||
|
||||
async def run(
|
||||
self,
|
||||
input_data: Input,
|
||||
*,
|
||||
credentials: DatabaseCredentials,
|
||||
**_kwargs: Any,
|
||||
) -> BlockOutput:
|
||||
# Validate query structure and read-only constraints.
|
||||
error = self._validate_query(input_data)
|
||||
if error:
|
||||
yield "error", error
|
||||
return
|
||||
|
||||
# Validate host and resolve for SSRF protection.
|
||||
host, pinned_host, error = await self._resolve_host(input_data)
|
||||
if error:
|
||||
yield "error", error
|
||||
return
|
||||
|
||||
# Build connection URL and execute.
|
||||
port = input_data.port or _DATABASE_TYPE_DEFAULT_PORT[input_data.database_type]
|
||||
username = credentials.username.get_secret_value()
|
||||
connection_url = URL.create(
|
||||
drivername=_DATABASE_TYPE_TO_DRIVER[input_data.database_type],
|
||||
username=username,
|
||||
password=credentials.password.get_secret_value(),
|
||||
host=pinned_host,
|
||||
port=port,
|
||||
database=input_data.database,
|
||||
)
|
||||
conn_str = connection_url.render_as_string(hide_password=True)
|
||||
db_name = input_data.database
|
||||
|
||||
def _sanitize(err: Exception) -> str:
|
||||
return _sanitize_error(
|
||||
str(err).strip(),
|
||||
conn_str,
|
||||
host=pinned_host,
|
||||
original_host=host,
|
||||
username=username,
|
||||
port=port,
|
||||
database=db_name,
|
||||
)
|
||||
|
||||
try:
|
||||
results, columns, affected = await asyncio.to_thread(
|
||||
self.execute_query,
|
||||
connection_url=connection_url,
|
||||
query=input_data.query,
|
||||
timeout=input_data.timeout,
|
||||
max_rows=input_data.max_rows,
|
||||
read_only=input_data.read_only,
|
||||
database_type=input_data.database_type,
|
||||
)
|
||||
yield "results", results
|
||||
yield "columns", columns
|
||||
yield "row_count", len(results)
|
||||
if affected >= 0:
|
||||
yield "affected_rows", affected
|
||||
except OperationalError as e:
|
||||
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."
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def _validate_query(input_data: "SQLQueryBlock.Input") -> str | None:
|
||||
"""Validate query structure and read-only constraints."""
|
||||
stmt_error, parsed_stmt = _validate_single_statement(input_data.query)
|
||||
if stmt_error:
|
||||
return stmt_error
|
||||
assert parsed_stmt is not None
|
||||
if input_data.read_only:
|
||||
return _validate_query_is_read_only(parsed_stmt)
|
||||
return None
|
||||
|
||||
async def _resolve_host(
|
||||
self, input_data: "SQLQueryBlock.Input"
|
||||
) -> tuple[str, str, str | None]:
|
||||
"""Validate and resolve the database host. Returns (host, pinned_ip, error)."""
|
||||
host = input_data.host.get_secret_value().strip()
|
||||
if not host:
|
||||
return "", "", "Database host is required."
|
||||
if host.startswith("/"):
|
||||
return host, "", "Unix socket connections are not allowed."
|
||||
try:
|
||||
resolved_ips = await self.check_host_allowed(host)
|
||||
except (ValueError, OSError) as e:
|
||||
return host, "", f"Blocked host: {str(e).strip()}"
|
||||
return host, resolved_ips[0], None
|
||||
|
||||
@staticmethod
|
||||
def _classify_operational_error(sanitized_msg: str, timeout: int) -> str:
|
||||
"""Classify an already-sanitized OperationalError for user display."""
|
||||
lower = sanitized_msg.lower()
|
||||
if "timeout" in lower or "cancel" in lower:
|
||||
return f"Query timed out after {timeout}s."
|
||||
if "connect" in lower:
|
||||
return f"Failed to connect to database: {sanitized_msg}"
|
||||
return f"Database error: {sanitized_msg}"
|
||||
File diff suppressed because it is too large
Load Diff
@@ -1,376 +0,0 @@
|
||||
import re
|
||||
from datetime import date, datetime, time
|
||||
from decimal import Decimal
|
||||
from enum import Enum
|
||||
from typing import Any
|
||||
|
||||
import sqlparse
|
||||
from sqlalchemy import create_engine, text
|
||||
from sqlalchemy.engine.url import URL
|
||||
|
||||
|
||||
class DatabaseType(str, Enum):
|
||||
POSTGRES = "postgres"
|
||||
MYSQL = "mysql"
|
||||
MSSQL = "mssql"
|
||||
|
||||
|
||||
# Defense-in-depth: reject queries containing data-modifying keywords.
|
||||
# These are checked against parsed SQL tokens (not raw text) so column names
|
||||
# and string literals do not cause false positives.
|
||||
_DISALLOWED_KEYWORDS = {
|
||||
"INSERT",
|
||||
"UPDATE",
|
||||
"DELETE",
|
||||
"DROP",
|
||||
"ALTER",
|
||||
"CREATE",
|
||||
"TRUNCATE",
|
||||
"GRANT",
|
||||
"REVOKE",
|
||||
"COPY",
|
||||
"EXECUTE",
|
||||
"CALL",
|
||||
"SET",
|
||||
"RESET",
|
||||
"DISCARD",
|
||||
"NOTIFY",
|
||||
"DO",
|
||||
}
|
||||
|
||||
# Map DatabaseType enum values to the expected SQLAlchemy driver prefix.
|
||||
_DATABASE_TYPE_TO_DRIVER = {
|
||||
DatabaseType.POSTGRES: "postgresql",
|
||||
DatabaseType.MYSQL: "mysql+pymysql",
|
||||
DatabaseType.MSSQL: "mssql+pymssql",
|
||||
}
|
||||
|
||||
# Default ports for each database type.
|
||||
_DATABASE_TYPE_DEFAULT_PORT = {
|
||||
DatabaseType.POSTGRES: 5432,
|
||||
DatabaseType.MYSQL: 3306,
|
||||
DatabaseType.MSSQL: 1433,
|
||||
}
|
||||
|
||||
|
||||
def _sanitize_error(
|
||||
error_msg: str,
|
||||
connection_string: str,
|
||||
*,
|
||||
host: str = "",
|
||||
original_host: str = "",
|
||||
username: str = "",
|
||||
port: int = 0,
|
||||
database: str = "",
|
||||
) -> str:
|
||||
"""Remove connection string, credentials, and infrastructure details
|
||||
from error messages so they are safe to expose to the LLM.
|
||||
|
||||
Scrubs:
|
||||
- The full connection string
|
||||
- URL-embedded credentials (``://user:pass@``)
|
||||
- ``password=<value>`` key-value pairs
|
||||
- The database hostname / IP used for the connection
|
||||
- The original (pre-resolution) hostname provided by the user
|
||||
- Any IPv4 addresses that appear in the message
|
||||
- Any bracketed IPv6 addresses (e.g. ``[::1]``, ``[fe80::1%eth0]``)
|
||||
- The database username
|
||||
- The database port number
|
||||
- The database name
|
||||
"""
|
||||
sanitized = error_msg.replace(connection_string, "<connection_string>")
|
||||
sanitized = re.sub(r"password=[^\s&]+", "password=***", sanitized)
|
||||
sanitized = re.sub(r"://[^@]+@", "://***:***@", sanitized)
|
||||
|
||||
# Replace the known host (may be an IP already) before the generic IP pass.
|
||||
# Also replace the original (pre-DNS-resolution) hostname if it differs.
|
||||
if original_host and original_host != host:
|
||||
sanitized = sanitized.replace(original_host, "<host>")
|
||||
if host:
|
||||
sanitized = sanitized.replace(host, "<host>")
|
||||
|
||||
# Replace any remaining IPv4 addresses (e.g. resolved IPs the driver logs)
|
||||
sanitized = re.sub(r"\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}", "<ip>", sanitized)
|
||||
|
||||
# Replace bracketed IPv6 addresses (e.g. "[::1]", "[fe80::1%eth0]")
|
||||
sanitized = re.sub(r"\[[0-9a-fA-F:]+(?:%[^\]]+)?\]", "<ip>", sanitized)
|
||||
|
||||
# Replace the database username (handles double-quoted, single-quoted,
|
||||
# and unquoted formats across PostgreSQL, MySQL, and MSSQL error messages).
|
||||
if username:
|
||||
sanitized = re.sub(
|
||||
r"""for user ["']?""" + re.escape(username) + r"""["']?""",
|
||||
"for user <user>",
|
||||
sanitized,
|
||||
)
|
||||
# Catch remaining bare occurrences in various quote styles:
|
||||
# - PostgreSQL: "FATAL: role "myuser" does not exist"
|
||||
# - MySQL: "Access denied for user 'myuser'@'host'"
|
||||
# - MSSQL: "Login failed for user 'myuser'"
|
||||
sanitized = sanitized.replace(f'"{username}"', "<user>")
|
||||
sanitized = sanitized.replace(f"'{username}'", "<user>")
|
||||
|
||||
# Replace the port number (handles "port 5432" and ":5432" formats)
|
||||
if port:
|
||||
port_str = re.escape(str(port))
|
||||
sanitized = re.sub(
|
||||
r"(?:port |:)" + port_str + r"(?![0-9])",
|
||||
lambda m: ("port " if m.group().startswith("p") else ":") + "<port>",
|
||||
sanitized,
|
||||
)
|
||||
|
||||
# Replace the database name to avoid leaking internal infrastructure names.
|
||||
# Use word-boundary regex to prevent mangling when the database name is a
|
||||
# common substring (e.g. "test", "data", "on").
|
||||
if database:
|
||||
sanitized = re.sub(r"\b" + re.escape(database) + r"\b", "<database>", sanitized)
|
||||
|
||||
return sanitized
|
||||
|
||||
|
||||
def _extract_keyword_tokens(parsed: sqlparse.sql.Statement) -> list[str]:
|
||||
"""Extract keyword tokens from a parsed SQL statement.
|
||||
|
||||
Uses sqlparse token type classification to collect Keyword/DML/DDL/DCL
|
||||
tokens. String literals and identifiers have different token types, so
|
||||
they are naturally excluded from the result.
|
||||
"""
|
||||
return [
|
||||
token.normalized.upper()
|
||||
for token in parsed.flatten()
|
||||
if token.ttype
|
||||
in (
|
||||
sqlparse.tokens.Keyword,
|
||||
sqlparse.tokens.Keyword.DML,
|
||||
sqlparse.tokens.Keyword.DDL,
|
||||
sqlparse.tokens.Keyword.DCL,
|
||||
)
|
||||
]
|
||||
|
||||
|
||||
def _has_disallowed_into(stmt: sqlparse.sql.Statement) -> bool:
|
||||
"""Check if a statement contains a disallowed ``INTO`` clause.
|
||||
|
||||
``SELECT ... INTO @variable`` is a valid read-only MySQL syntax that stores
|
||||
a query result into a session-scoped user variable. All other forms of
|
||||
``INTO`` are data-modifying or file-writing and must be blocked:
|
||||
|
||||
* ``SELECT ... INTO new_table`` (PostgreSQL / MSSQL – creates a table)
|
||||
* ``SELECT ... INTO OUTFILE`` (MySQL – writes to the filesystem)
|
||||
* ``SELECT ... INTO DUMPFILE`` (MySQL – writes to the filesystem)
|
||||
* ``INSERT INTO ...`` (already blocked by INSERT being in the
|
||||
disallowed set, but we reject INTO as well for defense-in-depth)
|
||||
|
||||
Returns ``True`` if the statement contains a disallowed ``INTO``.
|
||||
"""
|
||||
flat = list(stmt.flatten())
|
||||
for i, token in enumerate(flat):
|
||||
if not (
|
||||
token.ttype in (sqlparse.tokens.Keyword,)
|
||||
and token.normalized.upper() == "INTO"
|
||||
):
|
||||
continue
|
||||
|
||||
# Look at the first non-whitespace token after INTO.
|
||||
j = i + 1
|
||||
while j < len(flat) and flat[j].ttype is sqlparse.tokens.Text.Whitespace:
|
||||
j += 1
|
||||
|
||||
if j >= len(flat):
|
||||
# INTO at the very end – malformed, block it.
|
||||
return True
|
||||
|
||||
next_token = flat[j]
|
||||
# MySQL user variable: either a single Name starting with "@"
|
||||
# (e.g. ``@total``) or a bare ``@`` Operator token followed by a Name.
|
||||
if next_token.ttype is sqlparse.tokens.Name and next_token.value.startswith(
|
||||
"@"
|
||||
):
|
||||
continue
|
||||
if next_token.ttype is sqlparse.tokens.Operator and next_token.value == "@":
|
||||
continue
|
||||
|
||||
# Everything else (table name, OUTFILE, DUMPFILE, etc.) is disallowed.
|
||||
return True
|
||||
|
||||
return False
|
||||
|
||||
|
||||
def _validate_query_is_read_only(stmt: sqlparse.sql.Statement) -> str | None:
|
||||
"""Validate that a parsed SQL statement is read-only (SELECT/WITH only).
|
||||
|
||||
Accepts an already-parsed statement from ``_validate_single_statement``
|
||||
to avoid re-parsing. Checks:
|
||||
1. Statement type must be SELECT (sqlparse classifies WITH...SELECT as SELECT)
|
||||
2. No disallowed keywords (INSERT, UPDATE, DELETE, DROP, etc.)
|
||||
3. No disallowed INTO clauses (allows MySQL ``SELECT ... INTO @variable``)
|
||||
|
||||
Returns an error message if the query is not read-only, None otherwise.
|
||||
"""
|
||||
# sqlparse returns 'SELECT' for SELECT and WITH...SELECT queries
|
||||
if stmt.get_type() != "SELECT":
|
||||
return "Only SELECT queries are allowed."
|
||||
|
||||
# Defense-in-depth: check parsed keyword tokens for disallowed keywords
|
||||
for kw in _extract_keyword_tokens(stmt):
|
||||
# Normalize multi-word tokens (e.g. "SET LOCAL" -> "SET")
|
||||
base_kw = kw.split()[0] if " " in kw else kw
|
||||
if base_kw in _DISALLOWED_KEYWORDS:
|
||||
return f"Disallowed SQL keyword: {kw}"
|
||||
|
||||
# Contextual check for INTO: allow MySQL @variable syntax, block everything else
|
||||
if _has_disallowed_into(stmt):
|
||||
return "Disallowed SQL keyword: INTO"
|
||||
|
||||
return None
|
||||
|
||||
|
||||
def _validate_single_statement(
|
||||
query: str,
|
||||
) -> tuple[str | None, sqlparse.sql.Statement | None]:
|
||||
"""Validate that the query contains exactly one non-empty SQL statement.
|
||||
|
||||
Returns (error_message, parsed_statement). If error_message is not None,
|
||||
the query is invalid and parsed_statement will be None.
|
||||
"""
|
||||
stripped = query.strip().rstrip(";").strip()
|
||||
if not stripped:
|
||||
return "Query is empty.", None
|
||||
|
||||
# Parse the SQL using sqlparse for proper tokenization
|
||||
statements = sqlparse.parse(stripped)
|
||||
|
||||
# Filter out empty statements and comment-only statements
|
||||
statements = [
|
||||
s
|
||||
for s in statements
|
||||
if s.tokens
|
||||
and str(s).strip()
|
||||
and not all(
|
||||
t.is_whitespace or t.ttype in sqlparse.tokens.Comment for t in s.flatten()
|
||||
)
|
||||
]
|
||||
|
||||
if not statements:
|
||||
return "Query is empty.", None
|
||||
|
||||
# Reject multiple statements -- prevents injection via semicolons
|
||||
if len(statements) > 1:
|
||||
return "Only single statements are allowed.", None
|
||||
|
||||
return None, statements[0]
|
||||
|
||||
|
||||
def _serialize_value(value: Any) -> Any:
|
||||
"""Convert database-specific types to JSON-serializable Python types."""
|
||||
if isinstance(value, Decimal):
|
||||
# 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():
|
||||
return int(value)
|
||||
return str(value)
|
||||
if isinstance(value, (datetime, date, time)):
|
||||
return value.isoformat()
|
||||
if isinstance(value, memoryview):
|
||||
return bytes(value).hex()
|
||||
if isinstance(value, bytes):
|
||||
return value.hex()
|
||||
return value
|
||||
|
||||
|
||||
def _configure_session(
|
||||
conn: Any,
|
||||
dialect_name: str,
|
||||
timeout_ms: str,
|
||||
read_only: bool,
|
||||
) -> None:
|
||||
"""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:
|
||||
conn.execute(text("SET default_transaction_read_only = ON"))
|
||||
elif dialect_name == "mysql":
|
||||
# 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.
|
||||
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).
|
||||
# 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
|
||||
# the SQL validation layer (_validate_query_is_read_only)
|
||||
# and the ROLLBACK in the finally block.
|
||||
|
||||
|
||||
def _run_in_transaction(
|
||||
conn: Any,
|
||||
dialect_name: str,
|
||||
query: str,
|
||||
max_rows: int,
|
||||
read_only: bool,
|
||||
) -> 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))
|
||||
try:
|
||||
result = conn.execute(text(query))
|
||||
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 []
|
||||
results = [
|
||||
{col: _serialize_value(val) for col, val in zip(columns, row)}
|
||||
for row in rows
|
||||
]
|
||||
except Exception:
|
||||
conn.execute(text("ROLLBACK"))
|
||||
raise
|
||||
else:
|
||||
conn.execute(text("ROLLBACK" if read_only else "COMMIT"))
|
||||
return results, columns, affected
|
||||
|
||||
|
||||
def _execute_query(
|
||||
connection_url: URL | str,
|
||||
query: str,
|
||||
timeout: int,
|
||||
max_rows: int,
|
||||
read_only: bool = True,
|
||||
database_type: DatabaseType = DatabaseType.POSTGRES,
|
||||
) -> 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.
|
||||
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.
|
||||
"""
|
||||
# Determine driver-specific connection timeout argument.
|
||||
# pymssql uses "login_timeout", while PostgreSQL/MySQL use "connect_timeout".
|
||||
timeout_key = (
|
||||
"login_timeout" if database_type == DatabaseType.MSSQL else "connect_timeout"
|
||||
)
|
||||
engine = create_engine(connection_url, connect_args={timeout_key: 10})
|
||||
try:
|
||||
with engine.connect() as conn:
|
||||
# Use AUTOCOMMIT so SET commands take effect immediately.
|
||||
conn = conn.execution_options(isolation_level="AUTOCOMMIT")
|
||||
|
||||
# Compute timeout in milliseconds. The value is Pydantic-validated
|
||||
# (ge=1, le=120), but we use int() as defense-in-depth.
|
||||
# NOTE: SET commands do not support bind parameters in most
|
||||
# databases, so we use str(int(...)) for safe interpolation.
|
||||
timeout_ms = str(int(timeout * 1000))
|
||||
|
||||
_configure_session(conn, engine.dialect.name, timeout_ms, read_only)
|
||||
return _run_in_transaction(
|
||||
conn, engine.dialect.name, query, max_rows, read_only
|
||||
)
|
||||
finally:
|
||||
engine.dispose()
|
||||
@@ -15,7 +15,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -182,7 +181,6 @@ 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":
|
||||
|
||||
@@ -4,8 +4,6 @@ import pytest
|
||||
|
||||
from backend.blocks import get_blocks
|
||||
from backend.blocks._base import Block, BlockSchemaInput
|
||||
from backend.blocks.io import AgentDropdownInputBlock, AgentInputBlock
|
||||
from backend.data.graph import BaseGraph
|
||||
from backend.data.model import SchemaField
|
||||
from backend.util.test import execute_block_test
|
||||
|
||||
@@ -281,113 +279,3 @@ class TestAutoCredentialsFieldsValidation:
|
||||
assert "Duplicate auto_credentials kwarg_name 'credentials'" in str(
|
||||
exc_info.value
|
||||
)
|
||||
|
||||
|
||||
def test_agent_input_block_ignores_legacy_placeholder_values():
|
||||
"""Verify AgentInputBlock.Input.model_construct tolerates extra placeholder_values
|
||||
for backward compatibility with existing agent JSON."""
|
||||
legacy_data = {
|
||||
"name": "url",
|
||||
"value": "",
|
||||
"description": "Enter a URL",
|
||||
"placeholder_values": ["https://example.com"],
|
||||
}
|
||||
instance = AgentInputBlock.Input.model_construct(**legacy_data)
|
||||
schema = instance.generate_schema()
|
||||
assert (
|
||||
"enum" not in schema
|
||||
), "AgentInputBlock should not produce enum from legacy placeholder_values"
|
||||
|
||||
|
||||
def test_dropdown_input_block_produces_enum():
|
||||
"""Verify AgentDropdownInputBlock.Input.generate_schema() produces enum
|
||||
using the canonical 'options' field name."""
|
||||
opts = ["Option A", "Option B"]
|
||||
instance = AgentDropdownInputBlock.Input.model_construct(
|
||||
name="choice", value=None, options=opts
|
||||
)
|
||||
schema = instance.generate_schema()
|
||||
assert schema.get("enum") == opts
|
||||
|
||||
|
||||
def test_dropdown_input_block_legacy_placeholder_values_produces_enum():
|
||||
"""Verify backward compat: passing legacy 'placeholder_values' to
|
||||
AgentDropdownInputBlock still produces enum via model_construct remap."""
|
||||
opts = ["Option A", "Option B"]
|
||||
instance = AgentDropdownInputBlock.Input.model_construct(
|
||||
name="choice", value=None, placeholder_values=opts
|
||||
)
|
||||
schema = instance.generate_schema()
|
||||
assert (
|
||||
schema.get("enum") == opts
|
||||
), "Legacy placeholder_values should be remapped to options"
|
||||
|
||||
|
||||
def test_generate_schema_integration_legacy_placeholder_values():
|
||||
"""Test the full Graph._generate_schema path with legacy placeholder_values
|
||||
on AgentInputBlock — verifies no enum leaks through the graph loading path."""
|
||||
legacy_input_default = {
|
||||
"name": "url",
|
||||
"value": "",
|
||||
"description": "Enter a URL",
|
||||
"placeholder_values": ["https://example.com"],
|
||||
}
|
||||
result = BaseGraph._generate_schema(
|
||||
(AgentInputBlock.Input, legacy_input_default),
|
||||
)
|
||||
url_props = result["properties"]["url"]
|
||||
assert (
|
||||
"enum" not in url_props
|
||||
), "Graph schema should not contain enum from AgentInputBlock placeholder_values"
|
||||
|
||||
|
||||
def test_generate_schema_integration_dropdown_produces_enum():
|
||||
"""Test the full Graph._generate_schema path with AgentDropdownInputBlock
|
||||
— verifies enum IS produced for dropdown blocks using canonical field name."""
|
||||
dropdown_input_default = {
|
||||
"name": "color",
|
||||
"value": None,
|
||||
"options": ["Red", "Green", "Blue"],
|
||||
}
|
||||
result = BaseGraph._generate_schema(
|
||||
(AgentDropdownInputBlock.Input, dropdown_input_default),
|
||||
)
|
||||
color_props = result["properties"]["color"]
|
||||
assert color_props.get("enum") == [
|
||||
"Red",
|
||||
"Green",
|
||||
"Blue",
|
||||
], "Graph schema should contain enum from AgentDropdownInputBlock"
|
||||
|
||||
|
||||
def test_generate_schema_integration_dropdown_legacy_placeholder_values():
|
||||
"""Test the full Graph._generate_schema path with AgentDropdownInputBlock
|
||||
using legacy 'placeholder_values' — verifies backward compat produces enum."""
|
||||
legacy_dropdown_input_default = {
|
||||
"name": "color",
|
||||
"value": None,
|
||||
"placeholder_values": ["Red", "Green", "Blue"],
|
||||
}
|
||||
result = BaseGraph._generate_schema(
|
||||
(AgentDropdownInputBlock.Input, legacy_dropdown_input_default),
|
||||
)
|
||||
color_props = result["properties"]["color"]
|
||||
assert color_props.get("enum") == [
|
||||
"Red",
|
||||
"Green",
|
||||
"Blue",
|
||||
], "Legacy placeholder_values should still produce enum via model_construct remap"
|
||||
|
||||
|
||||
def test_dropdown_input_block_init_legacy_placeholder_values():
|
||||
"""Verify backward compat: constructing AgentDropdownInputBlock.Input via
|
||||
model_validate with legacy 'placeholder_values' correctly maps to 'options'."""
|
||||
opts = ["Option A", "Option B"]
|
||||
instance = AgentDropdownInputBlock.Input.model_validate(
|
||||
{"name": "choice", "value": None, "placeholder_values": opts}
|
||||
)
|
||||
assert (
|
||||
instance.options == opts
|
||||
), "Legacy placeholder_values should be remapped to options via model_validate"
|
||||
schema = instance.generate_schema()
|
||||
assert schema.get("enum") == opts
|
||||
|
||||
@@ -207,51 +207,6 @@ class TestXMLParserBlockSecurity:
|
||||
pass
|
||||
|
||||
|
||||
class TestXMLParserBlockSyntaxErrors:
|
||||
"""XML syntax errors should raise ValueError (not SyntaxError).
|
||||
|
||||
This ensures the base Block.execute() wraps them as BlockExecutionError
|
||||
(expected / user-caused) instead of BlockUnknownError (unexpected / alerts
|
||||
Sentry).
|
||||
"""
|
||||
|
||||
async def test_unclosed_tag_raises_value_error(self):
|
||||
"""Unclosed tags should raise ValueError, not SyntaxError."""
|
||||
block = XMLParserBlock()
|
||||
bad_xml = "<root><unclosed>"
|
||||
|
||||
with pytest.raises(ValueError, match="Unclosed tag"):
|
||||
async for _ in block.run(XMLParserBlock.Input(input_xml=bad_xml)):
|
||||
pass
|
||||
|
||||
async def test_unexpected_closing_tag_raises_value_error(self):
|
||||
"""Extra closing tags should raise ValueError, not SyntaxError."""
|
||||
block = XMLParserBlock()
|
||||
bad_xml = "</unexpected>"
|
||||
|
||||
with pytest.raises(ValueError):
|
||||
async for _ in block.run(XMLParserBlock.Input(input_xml=bad_xml)):
|
||||
pass
|
||||
|
||||
async def test_empty_xml_raises_value_error(self):
|
||||
"""Empty XML input should raise ValueError."""
|
||||
block = XMLParserBlock()
|
||||
|
||||
with pytest.raises(ValueError, match="XML input is empty"):
|
||||
async for _ in block.run(XMLParserBlock.Input(input_xml="")):
|
||||
pass
|
||||
|
||||
async def test_syntax_error_from_parser_becomes_value_error(self):
|
||||
"""SyntaxErrors from gravitasml library become ValueError (BlockExecutionError)."""
|
||||
block = XMLParserBlock()
|
||||
# Malformed XML that might trigger a SyntaxError from the parser
|
||||
bad_xml = "<root><child>no closing"
|
||||
|
||||
with pytest.raises(ValueError):
|
||||
async for _ in block.run(XMLParserBlock.Input(input_xml=bad_xml)):
|
||||
pass
|
||||
|
||||
|
||||
class TestStoreMediaFileSecurity:
|
||||
"""Test file storage security limits."""
|
||||
|
||||
|
||||
@@ -488,154 +488,6 @@ class TestLLMStatsTracking:
|
||||
assert outputs["response"] == {"result": "test"}
|
||||
|
||||
|
||||
class TestAIConversationBlockValidation:
|
||||
"""Test that AIConversationBlock validates inputs before calling the LLM."""
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_empty_messages_and_empty_prompt_raises_error(self):
|
||||
"""Empty messages with no prompt should raise ValueError, not a cryptic API error."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_empty_messages_with_prompt_succeeds(self):
|
||||
"""Empty messages but a non-empty prompt should proceed without error."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
async def mock_llm_call(input_data, credentials):
|
||||
return {"response": "OK"}
|
||||
|
||||
with patch.object(block, "llm_call", new=AsyncMock(side_effect=mock_llm_call)):
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[],
|
||||
prompt="Hello, how are you?",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
outputs = {}
|
||||
async for name, data in block.run(
|
||||
input_data, credentials=llm.TEST_CREDENTIALS
|
||||
):
|
||||
outputs[name] = data
|
||||
|
||||
assert outputs["response"] == "OK"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_nonempty_messages_with_empty_prompt_succeeds(self):
|
||||
"""Non-empty messages with no prompt should proceed without error."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
async def mock_llm_call(input_data, credentials):
|
||||
return {"response": "response from conversation"}
|
||||
|
||||
with patch.object(block, "llm_call", new=AsyncMock(side_effect=mock_llm_call)):
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[{"role": "user", "content": "Hello"}],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
outputs = {}
|
||||
async for name, data in block.run(
|
||||
input_data, credentials=llm.TEST_CREDENTIALS
|
||||
):
|
||||
outputs[name] = data
|
||||
|
||||
assert outputs["response"] == "response from conversation"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_messages_with_empty_content_raises_error(self):
|
||||
"""Messages with empty content strings should be treated as no messages."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[{"role": "user", "content": ""}],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_messages_with_whitespace_content_raises_error(self):
|
||||
"""Messages with whitespace-only content should be treated as no messages."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[{"role": "user", "content": " "}],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_messages_with_none_entry_raises_error(self):
|
||||
"""Messages list containing None should be treated as no messages."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[None],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_messages_with_empty_dict_raises_error(self):
|
||||
"""Messages list containing empty dict should be treated as no messages."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[{}],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_messages_with_none_content_raises_error(self):
|
||||
"""Messages with content=None should not crash with AttributeError."""
|
||||
block = llm.AIConversationBlock()
|
||||
|
||||
input_data = llm.AIConversationBlock.Input(
|
||||
messages=[{"role": "user", "content": None}],
|
||||
prompt="",
|
||||
model=llm.DEFAULT_LLM_MODEL,
|
||||
credentials=_TEST_AI_CREDENTIALS,
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="no messages and no prompt"):
|
||||
async for _ in block.run(input_data, credentials=llm.TEST_CREDENTIALS):
|
||||
pass
|
||||
|
||||
|
||||
class TestAITextSummarizerValidation:
|
||||
"""Test that AITextSummarizerBlock validates LLM responses are strings."""
|
||||
|
||||
@@ -957,33 +809,3 @@ class TestUserErrorStatusCodeHandling:
|
||||
|
||||
mock_warning.assert_called_once()
|
||||
mock_exception.assert_not_called()
|
||||
|
||||
|
||||
class TestLlmModelMissing:
|
||||
"""Test that LlmModel handles provider-prefixed model names."""
|
||||
|
||||
def test_provider_prefixed_model_resolves(self):
|
||||
"""Provider-prefixed model string should resolve to the correct enum member."""
|
||||
assert (
|
||||
llm.LlmModel("anthropic/claude-sonnet-4-6")
|
||||
== llm.LlmModel.CLAUDE_4_6_SONNET
|
||||
)
|
||||
|
||||
def test_bare_model_still_works(self):
|
||||
"""Bare (non-prefixed) model string should still resolve correctly."""
|
||||
assert llm.LlmModel("claude-sonnet-4-6") == llm.LlmModel.CLAUDE_4_6_SONNET
|
||||
|
||||
def test_invalid_prefixed_model_raises(self):
|
||||
"""Unknown provider-prefixed model string should raise ValueError."""
|
||||
with pytest.raises(ValueError):
|
||||
llm.LlmModel("invalid/nonexistent-model")
|
||||
|
||||
def test_slash_containing_value_direct_lookup(self):
|
||||
"""Enum values with '/' (e.g., OpenRouter models) should resolve via direct lookup, not _missing_."""
|
||||
assert llm.LlmModel("google/gemini-2.5-pro") == llm.LlmModel.GEMINI_2_5_PRO
|
||||
|
||||
def test_double_prefixed_slash_model(self):
|
||||
"""Double-prefixed value should still resolve by stripping first prefix."""
|
||||
assert (
|
||||
llm.LlmModel("extra/google/gemini-2.5-pro") == llm.LlmModel.GEMINI_2_5_PRO
|
||||
)
|
||||
|
||||
@@ -1,87 +0,0 @@
|
||||
"""Tests for empty-choices guard in extract_openai_tool_calls() and extract_openai_reasoning()."""
|
||||
|
||||
from unittest.mock import MagicMock
|
||||
|
||||
from backend.blocks.llm import extract_openai_reasoning, extract_openai_tool_calls
|
||||
|
||||
|
||||
class TestExtractOpenaiToolCallsEmptyChoices:
|
||||
"""extract_openai_tool_calls() must return None when choices is empty."""
|
||||
|
||||
def test_returns_none_for_empty_choices(self):
|
||||
response = MagicMock()
|
||||
response.choices = []
|
||||
assert extract_openai_tool_calls(response) is None
|
||||
|
||||
def test_returns_none_for_none_choices(self):
|
||||
response = MagicMock()
|
||||
response.choices = None
|
||||
assert extract_openai_tool_calls(response) is None
|
||||
|
||||
def test_returns_tool_calls_when_choices_present(self):
|
||||
tool = MagicMock()
|
||||
tool.id = "call_1"
|
||||
tool.type = "function"
|
||||
tool.function.name = "my_func"
|
||||
tool.function.arguments = '{"a": 1}'
|
||||
|
||||
message = MagicMock()
|
||||
message.tool_calls = [tool]
|
||||
|
||||
choice = MagicMock()
|
||||
choice.message = message
|
||||
|
||||
response = MagicMock()
|
||||
response.choices = [choice]
|
||||
|
||||
result = extract_openai_tool_calls(response)
|
||||
assert result is not None
|
||||
assert len(result) == 1
|
||||
assert result[0].function.name == "my_func"
|
||||
|
||||
def test_returns_none_when_no_tool_calls(self):
|
||||
message = MagicMock()
|
||||
message.tool_calls = None
|
||||
|
||||
choice = MagicMock()
|
||||
choice.message = message
|
||||
|
||||
response = MagicMock()
|
||||
response.choices = [choice]
|
||||
|
||||
assert extract_openai_tool_calls(response) is None
|
||||
|
||||
|
||||
class TestExtractOpenaiReasoningEmptyChoices:
|
||||
"""extract_openai_reasoning() must return None when choices is empty."""
|
||||
|
||||
def test_returns_none_for_empty_choices(self):
|
||||
response = MagicMock()
|
||||
response.choices = []
|
||||
assert extract_openai_reasoning(response) is None
|
||||
|
||||
def test_returns_none_for_none_choices(self):
|
||||
response = MagicMock()
|
||||
response.choices = None
|
||||
assert extract_openai_reasoning(response) is None
|
||||
|
||||
def test_returns_reasoning_from_choice(self):
|
||||
choice = MagicMock()
|
||||
choice.reasoning = "Step-by-step reasoning"
|
||||
choice.message = MagicMock(spec=[]) # no 'reasoning' attr on message
|
||||
|
||||
response = MagicMock(spec=[]) # no 'reasoning' attr on response
|
||||
response.choices = [choice]
|
||||
|
||||
result = extract_openai_reasoning(response)
|
||||
assert result == "Step-by-step reasoning"
|
||||
|
||||
def test_returns_none_when_no_reasoning(self):
|
||||
choice = MagicMock(spec=[]) # no 'reasoning' attr
|
||||
choice.message = MagicMock(spec=[]) # no 'reasoning' attr
|
||||
|
||||
response = MagicMock(spec=[]) # no 'reasoning' attr
|
||||
response.choices = [choice]
|
||||
|
||||
result = extract_openai_reasoning(response)
|
||||
assert result is None
|
||||
@@ -1074,7 +1074,6 @@ async def test_orchestrator_uses_customized_name_for_blocks():
|
||||
mock_node.block_id = StoreValueBlock().id
|
||||
mock_node.metadata = {"customized_name": "My Custom Tool Name"}
|
||||
mock_node.block = StoreValueBlock()
|
||||
mock_node.input_default = {}
|
||||
|
||||
# Create a mock link
|
||||
mock_link = MagicMock(spec=Link)
|
||||
@@ -1106,7 +1105,6 @@ async def test_orchestrator_falls_back_to_block_name():
|
||||
mock_node.block_id = StoreValueBlock().id
|
||||
mock_node.metadata = {} # No customized_name
|
||||
mock_node.block = StoreValueBlock()
|
||||
mock_node.input_default = {}
|
||||
|
||||
# Create a mock link
|
||||
mock_link = MagicMock(spec=Link)
|
||||
|
||||
@@ -1,202 +0,0 @@
|
||||
"""Tests for ExecutionMode enum and provider validation in the orchestrator.
|
||||
|
||||
Covers:
|
||||
- ExecutionMode enum members exist and have stable values
|
||||
- EXTENDED_THINKING provider validation (anthropic/open_router allowed, others rejected)
|
||||
- EXTENDED_THINKING model-name validation (must start with "claude")
|
||||
"""
|
||||
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
|
||||
import pytest
|
||||
|
||||
from backend.blocks.llm import LlmModel
|
||||
from backend.blocks.orchestrator import ExecutionMode, OrchestratorBlock
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# ExecutionMode enum integrity
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestExecutionModeEnum:
|
||||
"""Guard against accidental renames or removals of enum members."""
|
||||
|
||||
def test_built_in_exists(self):
|
||||
assert hasattr(ExecutionMode, "BUILT_IN")
|
||||
assert ExecutionMode.BUILT_IN.value == "built_in"
|
||||
|
||||
def test_extended_thinking_exists(self):
|
||||
assert hasattr(ExecutionMode, "EXTENDED_THINKING")
|
||||
assert ExecutionMode.EXTENDED_THINKING.value == "extended_thinking"
|
||||
|
||||
def test_exactly_two_members(self):
|
||||
"""If a new mode is added, this test should be updated intentionally."""
|
||||
assert set(ExecutionMode.__members__.keys()) == {
|
||||
"BUILT_IN",
|
||||
"EXTENDED_THINKING",
|
||||
}
|
||||
|
||||
def test_string_enum(self):
|
||||
"""ExecutionMode is a str enum so it serialises cleanly to JSON."""
|
||||
assert isinstance(ExecutionMode.BUILT_IN, str)
|
||||
assert isinstance(ExecutionMode.EXTENDED_THINKING, str)
|
||||
|
||||
def test_round_trip_from_value(self):
|
||||
"""Constructing from the string value should return the same member."""
|
||||
assert ExecutionMode("built_in") is ExecutionMode.BUILT_IN
|
||||
assert ExecutionMode("extended_thinking") is ExecutionMode.EXTENDED_THINKING
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Provider validation (inline in OrchestratorBlock.run)
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
def _make_model_stub(provider: str, value: str):
|
||||
"""Create a lightweight stub that behaves like LlmModel for validation."""
|
||||
metadata = MagicMock()
|
||||
metadata.provider = provider
|
||||
stub = MagicMock()
|
||||
stub.metadata = metadata
|
||||
stub.value = value
|
||||
return stub
|
||||
|
||||
|
||||
class TestExtendedThinkingProviderValidation:
|
||||
"""The orchestrator rejects EXTENDED_THINKING for non-Anthropic providers."""
|
||||
|
||||
def test_anthropic_provider_accepted(self):
|
||||
"""provider='anthropic' + claude model should not raise."""
|
||||
model = _make_model_stub("anthropic", "claude-opus-4-6")
|
||||
provider = model.metadata.provider
|
||||
model_name = model.value
|
||||
assert provider in ("anthropic", "open_router")
|
||||
assert model_name.startswith("claude")
|
||||
|
||||
def test_open_router_provider_accepted(self):
|
||||
"""provider='open_router' + claude model should not raise."""
|
||||
model = _make_model_stub("open_router", "claude-sonnet-4-6")
|
||||
provider = model.metadata.provider
|
||||
model_name = model.value
|
||||
assert provider in ("anthropic", "open_router")
|
||||
assert model_name.startswith("claude")
|
||||
|
||||
def test_openai_provider_rejected(self):
|
||||
"""provider='openai' should be rejected for EXTENDED_THINKING."""
|
||||
model = _make_model_stub("openai", "gpt-4o")
|
||||
provider = model.metadata.provider
|
||||
assert provider not in ("anthropic", "open_router")
|
||||
|
||||
def test_groq_provider_rejected(self):
|
||||
model = _make_model_stub("groq", "llama-3.3-70b-versatile")
|
||||
provider = model.metadata.provider
|
||||
assert provider not in ("anthropic", "open_router")
|
||||
|
||||
def test_non_claude_model_rejected_even_if_anthropic_provider(self):
|
||||
"""A hypothetical non-Claude model with provider='anthropic' is rejected."""
|
||||
model = _make_model_stub("anthropic", "not-a-claude-model")
|
||||
model_name = model.value
|
||||
assert not model_name.startswith("claude")
|
||||
|
||||
def test_real_gpt4o_model_rejected(self):
|
||||
"""Verify a real LlmModel enum member (GPT4O) fails the provider check."""
|
||||
model = LlmModel.GPT4O
|
||||
provider = model.metadata.provider
|
||||
assert provider not in ("anthropic", "open_router")
|
||||
|
||||
def test_real_claude_model_passes(self):
|
||||
"""Verify a real LlmModel enum member (CLAUDE_4_6_SONNET) passes."""
|
||||
model = LlmModel.CLAUDE_4_6_SONNET
|
||||
provider = model.metadata.provider
|
||||
model_name = model.value
|
||||
assert provider in ("anthropic", "open_router")
|
||||
assert model_name.startswith("claude")
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Integration-style: exercise the validation branch via OrchestratorBlock.run
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
def _make_input_data(model, execution_mode=ExecutionMode.EXTENDED_THINKING):
|
||||
"""Build a minimal MagicMock that satisfies OrchestratorBlock.run's early path."""
|
||||
inp = MagicMock()
|
||||
inp.execution_mode = execution_mode
|
||||
inp.model = model
|
||||
inp.prompt = "test"
|
||||
inp.sys_prompt = ""
|
||||
inp.conversation_history = []
|
||||
inp.last_tool_output = None
|
||||
inp.prompt_values = {}
|
||||
return inp
|
||||
|
||||
|
||||
async def _collect_run_outputs(block, input_data, **kwargs):
|
||||
"""Exhaust the OrchestratorBlock.run async generator, collecting outputs."""
|
||||
outputs = []
|
||||
async for item in block.run(input_data, **kwargs):
|
||||
outputs.append(item)
|
||||
return outputs
|
||||
|
||||
|
||||
class TestExtendedThinkingValidationRaisesInBlock:
|
||||
"""Call OrchestratorBlock.run far enough to trigger the ValueError."""
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_non_anthropic_provider_raises_valueerror(self):
|
||||
"""EXTENDED_THINKING + openai provider raises ValueError."""
|
||||
block = OrchestratorBlock()
|
||||
input_data = _make_input_data(model=LlmModel.GPT4O)
|
||||
|
||||
with (
|
||||
patch.object(
|
||||
block,
|
||||
"_create_tool_node_signatures",
|
||||
new_callable=AsyncMock,
|
||||
return_value=[],
|
||||
),
|
||||
pytest.raises(ValueError, match="Anthropic-compatible"),
|
||||
):
|
||||
await _collect_run_outputs(
|
||||
block,
|
||||
input_data,
|
||||
credentials=MagicMock(),
|
||||
graph_id="g",
|
||||
node_id="n",
|
||||
graph_exec_id="ge",
|
||||
node_exec_id="ne",
|
||||
user_id="u",
|
||||
graph_version=1,
|
||||
execution_context=MagicMock(),
|
||||
execution_processor=MagicMock(),
|
||||
)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_non_claude_model_with_anthropic_provider_raises(self):
|
||||
"""A model with anthropic provider but non-claude name raises ValueError."""
|
||||
block = OrchestratorBlock()
|
||||
fake_model = _make_model_stub("anthropic", "not-a-claude-model")
|
||||
input_data = _make_input_data(model=fake_model)
|
||||
|
||||
with (
|
||||
patch.object(
|
||||
block,
|
||||
"_create_tool_node_signatures",
|
||||
new_callable=AsyncMock,
|
||||
return_value=[],
|
||||
),
|
||||
pytest.raises(ValueError, match="only supports Claude models"),
|
||||
):
|
||||
await _collect_run_outputs(
|
||||
block,
|
||||
input_data,
|
||||
credentials=MagicMock(),
|
||||
graph_id="g",
|
||||
node_id="n",
|
||||
graph_exec_id="ge",
|
||||
node_exec_id="ne",
|
||||
user_id="u",
|
||||
graph_version=1,
|
||||
execution_context=MagicMock(),
|
||||
execution_processor=MagicMock(),
|
||||
)
|
||||
File diff suppressed because it is too large
Load Diff
@@ -13,7 +13,6 @@ from backend.data.model import (
|
||||
APIKeyCredentials,
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
)
|
||||
from backend.integrations.providers import ProviderName
|
||||
@@ -105,5 +104,4 @@ 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"]
|
||||
|
||||
@@ -44,7 +44,7 @@ class XMLParserBlock(Block):
|
||||
elif token.type == "TAG_CLOSE":
|
||||
depth -= 1
|
||||
if depth < 0:
|
||||
raise ValueError("Unexpected closing tag in XML input.")
|
||||
raise SyntaxError("Unexpected closing tag in XML input.")
|
||||
elif token.type in {"TEXT", "ESCAPE"}:
|
||||
if depth == 0 and token.value:
|
||||
raise ValueError(
|
||||
@@ -53,7 +53,7 @@ class XMLParserBlock(Block):
|
||||
)
|
||||
|
||||
if depth != 0:
|
||||
raise ValueError("Unclosed tag detected in XML input.")
|
||||
raise SyntaxError("Unclosed tag detected in XML input.")
|
||||
if not root_seen:
|
||||
raise ValueError("XML must include a root element.")
|
||||
|
||||
@@ -76,7 +76,4 @@ class XMLParserBlock(Block):
|
||||
except ValueError as val_e:
|
||||
raise ValueError(f"Validation error for dict:{val_e}") from val_e
|
||||
except SyntaxError as syn_e:
|
||||
# Raise as ValueError so the base Block.execute() wraps it as
|
||||
# BlockExecutionError (expected user-caused failure) instead of
|
||||
# BlockUnknownError (unexpected platform error that alerts Sentry).
|
||||
raise ValueError(f"Error in input xml syntax: {syn_e}") from syn_e
|
||||
raise SyntaxError(f"Error in input xml syntax: {syn_e}") from syn_e
|
||||
|
||||
@@ -19,7 +19,6 @@ from backend.blocks._base import (
|
||||
from backend.data.model import (
|
||||
CredentialsField,
|
||||
CredentialsMetaInput,
|
||||
NodeExecutionStats,
|
||||
SchemaField,
|
||||
UserPasswordCredentials,
|
||||
)
|
||||
@@ -171,7 +170,6 @@ 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
|
||||
|
||||
@@ -21,7 +21,7 @@ from backend.blocks.zerobounce._auth import (
|
||||
ZeroBounceCredentials,
|
||||
ZeroBounceCredentialsInput,
|
||||
)
|
||||
from backend.data.model import CredentialsField, NodeExecutionStats, SchemaField
|
||||
from backend.data.model import CredentialsField, SchemaField
|
||||
|
||||
|
||||
class Response(BaseModel):
|
||||
@@ -177,6 +177,5 @@ class ValidateEmailsBlock(Block):
|
||||
)
|
||||
|
||||
response_model = Response(**response.__dict__)
|
||||
self.merge_stats(NodeExecutionStats(output_size=1))
|
||||
|
||||
yield "response", response_model
|
||||
|
||||
@@ -9,16 +9,12 @@ shared tool registry as the SDK path.
|
||||
import asyncio
|
||||
import logging
|
||||
import uuid
|
||||
from collections.abc import AsyncGenerator, Sequence
|
||||
from dataclasses import dataclass, field
|
||||
from functools import partial
|
||||
from typing import Any, cast
|
||||
from collections.abc import AsyncGenerator
|
||||
from typing import Any
|
||||
|
||||
import orjson
|
||||
from langfuse import propagate_attributes
|
||||
from openai.types.chat import ChatCompletionMessageParam, ChatCompletionToolParam
|
||||
|
||||
from backend.copilot.context import set_execution_context
|
||||
from backend.copilot.model import (
|
||||
ChatMessage,
|
||||
ChatSession,
|
||||
@@ -51,24 +47,8 @@ from backend.copilot.service import (
|
||||
from backend.copilot.token_tracking import persist_and_record_usage
|
||||
from backend.copilot.tools import execute_tool, get_available_tools
|
||||
from backend.copilot.tracking import track_user_message
|
||||
from backend.copilot.transcript import (
|
||||
download_transcript,
|
||||
upload_transcript,
|
||||
validate_transcript,
|
||||
)
|
||||
from backend.copilot.transcript_builder import TranscriptBuilder
|
||||
from backend.util.exceptions import NotFoundError
|
||||
from backend.util.prompt import (
|
||||
compress_context,
|
||||
estimate_token_count,
|
||||
estimate_token_count_str,
|
||||
)
|
||||
from backend.util.tool_call_loop import (
|
||||
LLMLoopResponse,
|
||||
LLMToolCall,
|
||||
ToolCallResult,
|
||||
tool_call_loop,
|
||||
)
|
||||
from backend.util.prompt import compress_context
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@@ -79,286 +59,6 @@ _background_tasks: set[asyncio.Task[Any]] = set()
|
||||
_MAX_TOOL_ROUNDS = 30
|
||||
|
||||
|
||||
@dataclass
|
||||
class _BaselineStreamState:
|
||||
"""Mutable state shared between the tool-call loop callbacks.
|
||||
|
||||
Extracted from ``stream_chat_completion_baseline`` so that the callbacks
|
||||
can be module-level functions instead of deeply nested closures.
|
||||
"""
|
||||
|
||||
pending_events: list[StreamBaseResponse] = field(default_factory=list)
|
||||
assistant_text: str = ""
|
||||
text_block_id: str = field(default_factory=lambda: str(uuid.uuid4()))
|
||||
text_started: bool = False
|
||||
turn_prompt_tokens: int = 0
|
||||
turn_completion_tokens: int = 0
|
||||
|
||||
|
||||
async def _baseline_llm_caller(
|
||||
messages: list[dict[str, Any]],
|
||||
tools: Sequence[Any],
|
||||
*,
|
||||
state: _BaselineStreamState,
|
||||
) -> LLMLoopResponse:
|
||||
"""Stream an OpenAI-compatible response and collect results.
|
||||
|
||||
Extracted from ``stream_chat_completion_baseline`` for readability.
|
||||
"""
|
||||
state.pending_events.append(StreamStartStep())
|
||||
|
||||
round_text = ""
|
||||
try:
|
||||
client = _get_openai_client()
|
||||
typed_messages = cast(list[ChatCompletionMessageParam], messages)
|
||||
if tools:
|
||||
typed_tools = cast(list[ChatCompletionToolParam], tools)
|
||||
response = await client.chat.completions.create(
|
||||
model=config.fast_model,
|
||||
messages=typed_messages,
|
||||
tools=typed_tools,
|
||||
stream=True,
|
||||
stream_options={"include_usage": True},
|
||||
)
|
||||
else:
|
||||
response = await client.chat.completions.create(
|
||||
model=config.fast_model,
|
||||
messages=typed_messages,
|
||||
stream=True,
|
||||
stream_options={"include_usage": True},
|
||||
)
|
||||
tool_calls_by_index: dict[int, dict[str, str]] = {}
|
||||
|
||||
async for chunk in response:
|
||||
if chunk.usage:
|
||||
state.turn_prompt_tokens += chunk.usage.prompt_tokens or 0
|
||||
state.turn_completion_tokens += chunk.usage.completion_tokens or 0
|
||||
|
||||
delta = chunk.choices[0].delta if chunk.choices else None
|
||||
if not delta:
|
||||
continue
|
||||
|
||||
if delta.content:
|
||||
if not state.text_started:
|
||||
state.pending_events.append(StreamTextStart(id=state.text_block_id))
|
||||
state.text_started = True
|
||||
round_text += delta.content
|
||||
state.pending_events.append(
|
||||
StreamTextDelta(id=state.text_block_id, delta=delta.content)
|
||||
)
|
||||
|
||||
if delta.tool_calls:
|
||||
for tc in delta.tool_calls:
|
||||
idx = tc.index
|
||||
if idx not in tool_calls_by_index:
|
||||
tool_calls_by_index[idx] = {
|
||||
"id": "",
|
||||
"name": "",
|
||||
"arguments": "",
|
||||
}
|
||||
entry = tool_calls_by_index[idx]
|
||||
if tc.id:
|
||||
entry["id"] = tc.id
|
||||
if tc.function and tc.function.name:
|
||||
entry["name"] = tc.function.name
|
||||
if tc.function and tc.function.arguments:
|
||||
entry["arguments"] += tc.function.arguments
|
||||
|
||||
# Close text block
|
||||
if state.text_started:
|
||||
state.pending_events.append(StreamTextEnd(id=state.text_block_id))
|
||||
state.text_started = False
|
||||
state.text_block_id = str(uuid.uuid4())
|
||||
finally:
|
||||
# Always persist partial text so the session history stays consistent,
|
||||
# even when the stream is interrupted by an exception.
|
||||
state.assistant_text += round_text
|
||||
# Always emit StreamFinishStep to match the StreamStartStep,
|
||||
# even if an exception occurred during streaming.
|
||||
state.pending_events.append(StreamFinishStep())
|
||||
|
||||
# Convert to shared format
|
||||
llm_tool_calls = [
|
||||
LLMToolCall(
|
||||
id=tc["id"],
|
||||
name=tc["name"],
|
||||
arguments=tc["arguments"] or "{}",
|
||||
)
|
||||
for tc in tool_calls_by_index.values()
|
||||
]
|
||||
|
||||
return LLMLoopResponse(
|
||||
response_text=round_text or None,
|
||||
tool_calls=llm_tool_calls,
|
||||
raw_response=None, # Not needed for baseline conversation updater
|
||||
prompt_tokens=0, # Tracked via state accumulators
|
||||
completion_tokens=0,
|
||||
)
|
||||
|
||||
|
||||
async def _baseline_tool_executor(
|
||||
tool_call: LLMToolCall,
|
||||
tools: Sequence[Any],
|
||||
*,
|
||||
state: _BaselineStreamState,
|
||||
user_id: str | None,
|
||||
session: ChatSession,
|
||||
) -> ToolCallResult:
|
||||
"""Execute a tool via the copilot tool registry.
|
||||
|
||||
Extracted from ``stream_chat_completion_baseline`` for readability.
|
||||
"""
|
||||
tool_call_id = tool_call.id
|
||||
tool_name = tool_call.name
|
||||
raw_args = tool_call.arguments or "{}"
|
||||
|
||||
try:
|
||||
tool_args = orjson.loads(raw_args)
|
||||
except orjson.JSONDecodeError as parse_err:
|
||||
parse_error = f"Invalid JSON arguments for tool '{tool_name}': {parse_err}"
|
||||
logger.warning("[Baseline] %s", parse_error)
|
||||
state.pending_events.append(
|
||||
StreamToolOutputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
output=parse_error,
|
||||
success=False,
|
||||
)
|
||||
)
|
||||
return ToolCallResult(
|
||||
tool_call_id=tool_call_id,
|
||||
tool_name=tool_name,
|
||||
content=parse_error,
|
||||
is_error=True,
|
||||
)
|
||||
|
||||
state.pending_events.append(
|
||||
StreamToolInputStart(toolCallId=tool_call_id, toolName=tool_name)
|
||||
)
|
||||
state.pending_events.append(
|
||||
StreamToolInputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
input=tool_args,
|
||||
)
|
||||
)
|
||||
|
||||
try:
|
||||
result: StreamToolOutputAvailable = await execute_tool(
|
||||
tool_name=tool_name,
|
||||
parameters=tool_args,
|
||||
user_id=user_id,
|
||||
session=session,
|
||||
tool_call_id=tool_call_id,
|
||||
)
|
||||
state.pending_events.append(result)
|
||||
tool_output = (
|
||||
result.output if isinstance(result.output, str) else str(result.output)
|
||||
)
|
||||
return ToolCallResult(
|
||||
tool_call_id=tool_call_id,
|
||||
tool_name=tool_name,
|
||||
content=tool_output,
|
||||
)
|
||||
except Exception as e:
|
||||
error_output = f"Tool execution error: {e}"
|
||||
logger.error(
|
||||
"[Baseline] Tool %s failed: %s",
|
||||
tool_name,
|
||||
error_output,
|
||||
exc_info=True,
|
||||
)
|
||||
state.pending_events.append(
|
||||
StreamToolOutputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
output=error_output,
|
||||
success=False,
|
||||
)
|
||||
)
|
||||
return ToolCallResult(
|
||||
tool_call_id=tool_call_id,
|
||||
tool_name=tool_name,
|
||||
content=error_output,
|
||||
is_error=True,
|
||||
)
|
||||
|
||||
|
||||
def _baseline_conversation_updater(
|
||||
messages: list[dict[str, Any]],
|
||||
response: LLMLoopResponse,
|
||||
tool_results: list[ToolCallResult] | None = None,
|
||||
*,
|
||||
transcript_builder: TranscriptBuilder,
|
||||
model: str = "",
|
||||
) -> None:
|
||||
"""Update OpenAI message list with assistant response + tool results.
|
||||
|
||||
Extracted from ``stream_chat_completion_baseline`` for readability.
|
||||
"""
|
||||
if tool_results:
|
||||
# Build assistant message with tool_calls
|
||||
assistant_msg: dict[str, Any] = {"role": "assistant"}
|
||||
if response.response_text:
|
||||
assistant_msg["content"] = response.response_text
|
||||
assistant_msg["tool_calls"] = [
|
||||
{
|
||||
"id": tc.id,
|
||||
"type": "function",
|
||||
"function": {"name": tc.name, "arguments": tc.arguments},
|
||||
}
|
||||
for tc in response.tool_calls
|
||||
]
|
||||
messages.append(assistant_msg)
|
||||
# Record assistant message (with tool_calls) to transcript
|
||||
content_blocks: list[dict[str, Any]] = []
|
||||
if response.response_text:
|
||||
content_blocks.append({"type": "text", "text": response.response_text})
|
||||
for tc in response.tool_calls:
|
||||
try:
|
||||
args = orjson.loads(tc.arguments) if tc.arguments else {}
|
||||
except Exception:
|
||||
args = {}
|
||||
content_blocks.append(
|
||||
{
|
||||
"type": "tool_use",
|
||||
"id": tc.id,
|
||||
"name": tc.name,
|
||||
"input": args,
|
||||
}
|
||||
)
|
||||
if content_blocks:
|
||||
transcript_builder.append_assistant(
|
||||
content_blocks=content_blocks,
|
||||
model=model,
|
||||
stop_reason="tool_use",
|
||||
)
|
||||
for tr in tool_results:
|
||||
messages.append(
|
||||
{
|
||||
"role": "tool",
|
||||
"tool_call_id": tr.tool_call_id,
|
||||
"content": tr.content,
|
||||
}
|
||||
)
|
||||
# Record tool result to transcript AFTER the assistant tool_use
|
||||
# block to maintain correct Anthropic API ordering:
|
||||
# assistant(tool_use) → user(tool_result)
|
||||
transcript_builder.append_tool_result(
|
||||
tool_use_id=tr.tool_call_id,
|
||||
content=tr.content,
|
||||
)
|
||||
else:
|
||||
if response.response_text:
|
||||
messages.append({"role": "assistant", "content": response.response_text})
|
||||
# Record final text to transcript
|
||||
transcript_builder.append_assistant(
|
||||
content_blocks=[{"type": "text", "text": response.response_text}],
|
||||
model=model,
|
||||
stop_reason="end_turn",
|
||||
)
|
||||
|
||||
|
||||
async def _update_title_async(
|
||||
session_id: str, message: str, user_id: str | None
|
||||
) -> None:
|
||||
@@ -385,23 +85,19 @@ async def _compress_session_messages(
|
||||
msg_dict: dict[str, Any] = {"role": msg.role}
|
||||
if msg.content:
|
||||
msg_dict["content"] = msg.content
|
||||
if msg.tool_calls:
|
||||
msg_dict["tool_calls"] = msg.tool_calls
|
||||
if msg.tool_call_id:
|
||||
msg_dict["tool_call_id"] = msg.tool_call_id
|
||||
messages_dict.append(msg_dict)
|
||||
|
||||
try:
|
||||
result = await compress_context(
|
||||
messages=messages_dict,
|
||||
model=config.fast_model,
|
||||
model=config.model,
|
||||
client=_get_openai_client(),
|
||||
)
|
||||
except Exception as e:
|
||||
logger.warning("[Baseline] Context compression with LLM failed: %s", e)
|
||||
result = await compress_context(
|
||||
messages=messages_dict,
|
||||
model=config.fast_model,
|
||||
model=config.model,
|
||||
client=None,
|
||||
)
|
||||
|
||||
@@ -415,12 +111,7 @@ async def _compress_session_messages(
|
||||
result.messages_dropped,
|
||||
)
|
||||
return [
|
||||
ChatMessage(
|
||||
role=m["role"],
|
||||
content=m.get("content"),
|
||||
tool_calls=m.get("tool_calls"),
|
||||
tool_call_id=m.get("tool_call_id"),
|
||||
)
|
||||
ChatMessage(role=m["role"], content=m.get("content"))
|
||||
for m in result.messages
|
||||
]
|
||||
|
||||
@@ -451,8 +142,7 @@ async def stream_chat_completion_baseline(
|
||||
f"Session {session_id} not found. Please create a new session first."
|
||||
)
|
||||
|
||||
# Append user message (skip if it's an exact duplicate of the last message,
|
||||
# e.g. from a network retry)
|
||||
# Append user message
|
||||
new_role = "user" if is_user_message else "assistant"
|
||||
if message and (
|
||||
len(session.messages) == 0
|
||||
@@ -471,54 +161,6 @@ async def stream_chat_completion_baseline(
|
||||
|
||||
session = await upsert_chat_session(session)
|
||||
|
||||
# --- Transcript support (feature parity with SDK path) ---
|
||||
transcript_builder = TranscriptBuilder()
|
||||
transcript_covers_prefix = True
|
||||
|
||||
if user_id and len(session.messages) > 1:
|
||||
try:
|
||||
dl = await download_transcript(user_id, session_id, log_prefix="[Baseline]")
|
||||
if dl and validate_transcript(dl.content):
|
||||
# Reject stale transcripts: if msg_count is known and
|
||||
# doesn't cover the current session, loading it would
|
||||
# silently drop intermediate turns from the transcript.
|
||||
session_msg_count = len(session.messages)
|
||||
if dl.message_count and dl.message_count < session_msg_count - 1:
|
||||
logger.warning(
|
||||
"[Baseline] Transcript stale: covers %d of %d messages, skipping",
|
||||
dl.message_count,
|
||||
session_msg_count,
|
||||
)
|
||||
transcript_covers_prefix = False
|
||||
else:
|
||||
transcript_builder.load_previous(
|
||||
dl.content, log_prefix="[Baseline]"
|
||||
)
|
||||
logger.info(
|
||||
"[Baseline] Loaded transcript: %dB, msg_count=%d",
|
||||
len(dl.content),
|
||||
dl.message_count,
|
||||
)
|
||||
elif dl:
|
||||
logger.warning("[Baseline] Downloaded transcript but invalid")
|
||||
transcript_covers_prefix = False
|
||||
else:
|
||||
logger.debug("[Baseline] No transcript available")
|
||||
transcript_covers_prefix = False
|
||||
except Exception as e:
|
||||
logger.warning("[Baseline] Transcript download failed: %s", e)
|
||||
transcript_covers_prefix = False
|
||||
|
||||
# Append user message to transcript.
|
||||
# Always append when the message is present and is from the user,
|
||||
# even on duplicate-suppressed retries (is_new_message=False).
|
||||
# The loaded transcript may be stale (uploaded before the previous
|
||||
# attempt stored this message), so skipping it would leave the
|
||||
# transcript without the user turn, creating a malformed
|
||||
# assistant-after-assistant structure when the LLM reply is added.
|
||||
if message and is_user_message:
|
||||
transcript_builder.append_user(content=message)
|
||||
|
||||
# Generate title for new sessions
|
||||
if is_user_message and not session.title:
|
||||
user_messages = [m for m in session.messages if m.role == "user"]
|
||||
@@ -551,37 +193,16 @@ async def stream_chat_completion_baseline(
|
||||
# Compress context if approaching the model's token limit
|
||||
messages_for_context = await _compress_session_messages(session.messages)
|
||||
|
||||
# Build OpenAI message list from session history.
|
||||
# Include tool_calls on assistant messages and tool-role results so the
|
||||
# model retains full context of what tools were invoked and their outcomes.
|
||||
# Build OpenAI message list from session history
|
||||
openai_messages: list[dict[str, Any]] = [
|
||||
{"role": "system", "content": system_prompt}
|
||||
]
|
||||
for msg in messages_for_context:
|
||||
if msg.role == "assistant":
|
||||
entry: dict[str, Any] = {"role": "assistant"}
|
||||
if msg.content:
|
||||
entry["content"] = msg.content
|
||||
if msg.tool_calls:
|
||||
entry["tool_calls"] = msg.tool_calls
|
||||
if msg.content or msg.tool_calls:
|
||||
openai_messages.append(entry)
|
||||
elif msg.role == "tool" and msg.tool_call_id:
|
||||
openai_messages.append(
|
||||
{
|
||||
"role": "tool",
|
||||
"tool_call_id": msg.tool_call_id,
|
||||
"content": msg.content or "",
|
||||
}
|
||||
)
|
||||
elif msg.role == "user" and msg.content:
|
||||
if msg.role in ("user", "assistant") and msg.content:
|
||||
openai_messages.append({"role": msg.role, "content": msg.content})
|
||||
|
||||
tools = get_available_tools()
|
||||
|
||||
# Propagate execution context so tool handlers can read session-level flags.
|
||||
set_execution_context(user_id, session)
|
||||
|
||||
yield StreamStart(messageId=message_id, sessionId=session_id)
|
||||
|
||||
# Propagate user/session context to Langfuse so all LLM calls within
|
||||
@@ -598,38 +219,191 @@ async def stream_chat_completion_baseline(
|
||||
except Exception:
|
||||
logger.warning("[Baseline] Langfuse trace context setup failed")
|
||||
|
||||
assistant_text = ""
|
||||
text_block_id = str(uuid.uuid4())
|
||||
text_started = False
|
||||
step_open = False
|
||||
# Token usage accumulators — populated from streaming chunks
|
||||
turn_prompt_tokens = 0
|
||||
turn_completion_tokens = 0
|
||||
_stream_error = False # Track whether an error occurred during streaming
|
||||
state = _BaselineStreamState()
|
||||
|
||||
# Bind extracted module-level callbacks to this request's state/session
|
||||
# using functools.partial so they satisfy the Protocol signatures.
|
||||
_bound_llm_caller = partial(_baseline_llm_caller, state=state)
|
||||
_bound_tool_executor = partial(
|
||||
_baseline_tool_executor, state=state, user_id=user_id, session=session
|
||||
)
|
||||
|
||||
_bound_conversation_updater = partial(
|
||||
_baseline_conversation_updater,
|
||||
transcript_builder=transcript_builder,
|
||||
model=config.fast_model,
|
||||
)
|
||||
|
||||
try:
|
||||
loop_result = None
|
||||
async for loop_result in tool_call_loop(
|
||||
messages=openai_messages,
|
||||
tools=tools,
|
||||
llm_call=_bound_llm_caller,
|
||||
execute_tool=_bound_tool_executor,
|
||||
update_conversation=_bound_conversation_updater,
|
||||
max_iterations=_MAX_TOOL_ROUNDS,
|
||||
):
|
||||
# Drain buffered events after each iteration (real-time streaming)
|
||||
for evt in state.pending_events:
|
||||
yield evt
|
||||
state.pending_events.clear()
|
||||
for _round in range(_MAX_TOOL_ROUNDS):
|
||||
# Open a new step for each LLM round
|
||||
yield StreamStartStep()
|
||||
step_open = True
|
||||
|
||||
if loop_result and not loop_result.finished_naturally:
|
||||
# Stream a response from the model
|
||||
create_kwargs: dict[str, Any] = dict(
|
||||
model=config.model,
|
||||
messages=openai_messages,
|
||||
stream=True,
|
||||
stream_options={"include_usage": True},
|
||||
)
|
||||
if tools:
|
||||
create_kwargs["tools"] = tools
|
||||
response = await _get_openai_client().chat.completions.create(**create_kwargs) # type: ignore[arg-type] # dynamic kwargs
|
||||
|
||||
# Accumulate streamed response (text + tool calls)
|
||||
round_text = ""
|
||||
tool_calls_by_index: dict[int, dict[str, str]] = {}
|
||||
|
||||
async for chunk in response:
|
||||
# Capture token usage from the streaming chunk.
|
||||
# OpenRouter normalises all providers into OpenAI format
|
||||
# where prompt_tokens already includes cached tokens
|
||||
# (unlike Anthropic's native API). Use += to sum all
|
||||
# tool-call rounds since each API call is independent.
|
||||
# NOTE: stream_options={"include_usage": True} is not
|
||||
# universally supported — some providers (Mistral, Llama
|
||||
# via OpenRouter) always return chunk.usage=None. When
|
||||
# that happens, tokens stay 0 and the tiktoken fallback
|
||||
# below activates. Fail-open: one round is estimated.
|
||||
if chunk.usage:
|
||||
turn_prompt_tokens += chunk.usage.prompt_tokens or 0
|
||||
turn_completion_tokens += chunk.usage.completion_tokens or 0
|
||||
|
||||
delta = chunk.choices[0].delta if chunk.choices else None
|
||||
if not delta:
|
||||
continue
|
||||
|
||||
# Text content
|
||||
if delta.content:
|
||||
if not text_started:
|
||||
yield StreamTextStart(id=text_block_id)
|
||||
text_started = True
|
||||
round_text += delta.content
|
||||
yield StreamTextDelta(id=text_block_id, delta=delta.content)
|
||||
|
||||
# Tool call fragments (streamed incrementally)
|
||||
if delta.tool_calls:
|
||||
for tc in delta.tool_calls:
|
||||
idx = tc.index
|
||||
if idx not in tool_calls_by_index:
|
||||
tool_calls_by_index[idx] = {
|
||||
"id": "",
|
||||
"name": "",
|
||||
"arguments": "",
|
||||
}
|
||||
entry = tool_calls_by_index[idx]
|
||||
if tc.id:
|
||||
entry["id"] = tc.id
|
||||
if tc.function and tc.function.name:
|
||||
entry["name"] = tc.function.name
|
||||
if tc.function and tc.function.arguments:
|
||||
entry["arguments"] += tc.function.arguments
|
||||
|
||||
# Close text block if we had one this round
|
||||
if text_started:
|
||||
yield StreamTextEnd(id=text_block_id)
|
||||
text_started = False
|
||||
text_block_id = str(uuid.uuid4())
|
||||
|
||||
# Accumulate text for session persistence
|
||||
assistant_text += round_text
|
||||
|
||||
# No tool calls -> model is done
|
||||
if not tool_calls_by_index:
|
||||
yield StreamFinishStep()
|
||||
step_open = False
|
||||
break
|
||||
|
||||
# Close step before tool execution
|
||||
yield StreamFinishStep()
|
||||
step_open = False
|
||||
|
||||
# Append the assistant message with tool_calls to context.
|
||||
assistant_msg: dict[str, Any] = {"role": "assistant"}
|
||||
if round_text:
|
||||
assistant_msg["content"] = round_text
|
||||
assistant_msg["tool_calls"] = [
|
||||
{
|
||||
"id": tc["id"],
|
||||
"type": "function",
|
||||
"function": {
|
||||
"name": tc["name"],
|
||||
"arguments": tc["arguments"] or "{}",
|
||||
},
|
||||
}
|
||||
for tc in tool_calls_by_index.values()
|
||||
]
|
||||
openai_messages.append(assistant_msg)
|
||||
|
||||
# Execute each tool call and stream events
|
||||
for tc in tool_calls_by_index.values():
|
||||
tool_call_id = tc["id"]
|
||||
tool_name = tc["name"]
|
||||
raw_args = tc["arguments"] or "{}"
|
||||
try:
|
||||
tool_args = orjson.loads(raw_args)
|
||||
except orjson.JSONDecodeError as parse_err:
|
||||
parse_error = (
|
||||
f"Invalid JSON arguments for tool '{tool_name}': {parse_err}"
|
||||
)
|
||||
logger.warning("[Baseline] %s", parse_error)
|
||||
yield StreamToolOutputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
output=parse_error,
|
||||
success=False,
|
||||
)
|
||||
openai_messages.append(
|
||||
{
|
||||
"role": "tool",
|
||||
"tool_call_id": tool_call_id,
|
||||
"content": parse_error,
|
||||
}
|
||||
)
|
||||
continue
|
||||
|
||||
yield StreamToolInputStart(toolCallId=tool_call_id, toolName=tool_name)
|
||||
yield StreamToolInputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
input=tool_args,
|
||||
)
|
||||
|
||||
# Execute via shared tool registry
|
||||
try:
|
||||
result: StreamToolOutputAvailable = await execute_tool(
|
||||
tool_name=tool_name,
|
||||
parameters=tool_args,
|
||||
user_id=user_id,
|
||||
session=session,
|
||||
tool_call_id=tool_call_id,
|
||||
)
|
||||
yield result
|
||||
tool_output = (
|
||||
result.output
|
||||
if isinstance(result.output, str)
|
||||
else str(result.output)
|
||||
)
|
||||
except Exception as e:
|
||||
error_output = f"Tool execution error: {e}"
|
||||
logger.error(
|
||||
"[Baseline] Tool %s failed: %s",
|
||||
tool_name,
|
||||
error_output,
|
||||
exc_info=True,
|
||||
)
|
||||
yield StreamToolOutputAvailable(
|
||||
toolCallId=tool_call_id,
|
||||
toolName=tool_name,
|
||||
output=error_output,
|
||||
success=False,
|
||||
)
|
||||
tool_output = error_output
|
||||
|
||||
# Append tool result to context for next round
|
||||
openai_messages.append(
|
||||
{
|
||||
"role": "tool",
|
||||
"tool_call_id": tool_call_id,
|
||||
"content": tool_output,
|
||||
}
|
||||
)
|
||||
else:
|
||||
# for-loop exhausted without break -> tool-round limit hit
|
||||
limit_msg = (
|
||||
f"Exceeded {_MAX_TOOL_ROUNDS} tool-call rounds "
|
||||
"without a final response."
|
||||
@@ -644,28 +418,11 @@ async def stream_chat_completion_baseline(
|
||||
_stream_error = True
|
||||
error_msg = str(e) or type(e).__name__
|
||||
logger.error("[Baseline] Streaming error: %s", error_msg, exc_info=True)
|
||||
# Close any open text block. The llm_caller's finally block
|
||||
# already appended StreamFinishStep to pending_events, so we must
|
||||
# insert StreamTextEnd *before* StreamFinishStep to preserve the
|
||||
# protocol ordering:
|
||||
# StreamStartStep -> StreamTextStart -> ...deltas... ->
|
||||
# StreamTextEnd -> StreamFinishStep
|
||||
# Appending (or yielding directly) would place it after
|
||||
# StreamFinishStep, violating the protocol.
|
||||
if state.text_started:
|
||||
# Find the last StreamFinishStep and insert before it.
|
||||
insert_pos = len(state.pending_events)
|
||||
for i in range(len(state.pending_events) - 1, -1, -1):
|
||||
if isinstance(state.pending_events[i], StreamFinishStep):
|
||||
insert_pos = i
|
||||
break
|
||||
state.pending_events.insert(
|
||||
insert_pos, StreamTextEnd(id=state.text_block_id)
|
||||
)
|
||||
# Drain pending events in correct order
|
||||
for evt in state.pending_events:
|
||||
yield evt
|
||||
state.pending_events.clear()
|
||||
# Close any open text/step before emitting error
|
||||
if text_started:
|
||||
yield StreamTextEnd(id=text_block_id)
|
||||
if step_open:
|
||||
yield StreamFinishStep()
|
||||
yield StreamError(errorText=error_msg, code="baseline_error")
|
||||
# Still persist whatever we got
|
||||
finally:
|
||||
@@ -685,21 +442,26 @@ async def stream_chat_completion_baseline(
|
||||
# Skip fallback when an error occurred and no output was produced —
|
||||
# charging rate-limit tokens for completely failed requests is unfair.
|
||||
if (
|
||||
state.turn_prompt_tokens == 0
|
||||
and state.turn_completion_tokens == 0
|
||||
and not (_stream_error and not state.assistant_text)
|
||||
turn_prompt_tokens == 0
|
||||
and turn_completion_tokens == 0
|
||||
and not (_stream_error and not assistant_text)
|
||||
):
|
||||
state.turn_prompt_tokens = max(
|
||||
estimate_token_count(openai_messages, model=config.fast_model), 1
|
||||
from backend.util.prompt import (
|
||||
estimate_token_count,
|
||||
estimate_token_count_str,
|
||||
)
|
||||
state.turn_completion_tokens = estimate_token_count_str(
|
||||
state.assistant_text, model=config.fast_model
|
||||
|
||||
turn_prompt_tokens = max(
|
||||
estimate_token_count(openai_messages, model=config.model), 1
|
||||
)
|
||||
turn_completion_tokens = estimate_token_count_str(
|
||||
assistant_text, model=config.model
|
||||
)
|
||||
logger.info(
|
||||
"[Baseline] No streaming usage reported; estimated tokens: "
|
||||
"prompt=%d, completion=%d",
|
||||
state.turn_prompt_tokens,
|
||||
state.turn_completion_tokens,
|
||||
turn_prompt_tokens,
|
||||
turn_completion_tokens,
|
||||
)
|
||||
|
||||
# Persist token usage to session and record for rate limiting.
|
||||
@@ -709,50 +471,31 @@ async def stream_chat_completion_baseline(
|
||||
await persist_and_record_usage(
|
||||
session=session,
|
||||
user_id=user_id,
|
||||
prompt_tokens=state.turn_prompt_tokens,
|
||||
completion_tokens=state.turn_completion_tokens,
|
||||
prompt_tokens=turn_prompt_tokens,
|
||||
completion_tokens=turn_completion_tokens,
|
||||
log_prefix="[Baseline]",
|
||||
)
|
||||
|
||||
# Persist assistant response
|
||||
if state.assistant_text:
|
||||
if assistant_text:
|
||||
session.messages.append(
|
||||
ChatMessage(role="assistant", content=state.assistant_text)
|
||||
ChatMessage(role="assistant", content=assistant_text)
|
||||
)
|
||||
try:
|
||||
await upsert_chat_session(session)
|
||||
except Exception as persist_err:
|
||||
logger.error("[Baseline] Failed to persist session: %s", persist_err)
|
||||
|
||||
# --- Upload transcript for next-turn continuity ---
|
||||
if user_id and transcript_covers_prefix:
|
||||
try:
|
||||
_transcript_content = transcript_builder.to_jsonl()
|
||||
if _transcript_content and validate_transcript(_transcript_content):
|
||||
await asyncio.shield(
|
||||
upload_transcript(
|
||||
user_id=user_id,
|
||||
session_id=session_id,
|
||||
content=_transcript_content,
|
||||
message_count=len(session.messages),
|
||||
log_prefix="[Baseline]",
|
||||
)
|
||||
)
|
||||
else:
|
||||
logger.debug("[Baseline] No valid transcript to upload")
|
||||
except Exception as upload_err:
|
||||
logger.error("[Baseline] Transcript upload failed: %s", upload_err)
|
||||
|
||||
# Yield usage and finish AFTER try/finally (not inside finally).
|
||||
# PEP 525 prohibits yielding from finally in async generators during
|
||||
# aclose() — doing so raises RuntimeError on client disconnect.
|
||||
# On GeneratorExit the client is already gone, so unreachable yields
|
||||
# are harmless; on normal completion they reach the SSE stream.
|
||||
if state.turn_prompt_tokens > 0 or state.turn_completion_tokens > 0:
|
||||
if turn_prompt_tokens > 0 or turn_completion_tokens > 0:
|
||||
yield StreamUsage(
|
||||
prompt_tokens=state.turn_prompt_tokens,
|
||||
completion_tokens=state.turn_completion_tokens,
|
||||
total_tokens=state.turn_prompt_tokens + state.turn_completion_tokens,
|
||||
prompt_tokens=turn_prompt_tokens,
|
||||
completion_tokens=turn_completion_tokens,
|
||||
total_tokens=turn_prompt_tokens + turn_completion_tokens,
|
||||
)
|
||||
|
||||
yield StreamFinish()
|
||||
|
||||
@@ -31,7 +31,7 @@ async def test_baseline_multi_turn(setup_test_user, test_user_id):
|
||||
if not api_key:
|
||||
return pytest.skip("OPEN_ROUTER_API_KEY is not set, skipping test")
|
||||
|
||||
session = await create_chat_session(test_user_id, dry_run=False)
|
||||
session = await create_chat_session(test_user_id)
|
||||
session = await upsert_chat_session(session)
|
||||
|
||||
# --- Turn 1: send a message with a unique keyword ---
|
||||
|
||||
@@ -14,21 +14,12 @@ class ChatConfig(BaseSettings):
|
||||
|
||||
# OpenAI API Configuration
|
||||
model: str = Field(
|
||||
default="anthropic/claude-opus-4.6",
|
||||
description="Default model for extended thinking mode",
|
||||
)
|
||||
fast_model: str = Field(
|
||||
default="anthropic/claude-sonnet-4",
|
||||
description="Model for fast mode (baseline path). Should be faster/cheaper than the default model.",
|
||||
default="anthropic/claude-opus-4.6", description="Default model to use"
|
||||
)
|
||||
title_model: str = Field(
|
||||
default="openai/gpt-4o-mini",
|
||||
description="Model to use for generating session titles (should be fast/cheap)",
|
||||
)
|
||||
simulation_model: str = Field(
|
||||
default="google/gemini-2.5-flash",
|
||||
description="Model for dry-run block simulation (should be fast/cheap with good JSON output)",
|
||||
)
|
||||
api_key: str | None = Field(default=None, description="OpenAI API key")
|
||||
base_url: str | None = Field(
|
||||
default=OPENROUTER_BASE_URL,
|
||||
@@ -86,11 +77,11 @@ class ChatConfig(BaseSettings):
|
||||
# allows ~70-100 turns/day.
|
||||
# Checked at the HTTP layer (routes.py) before each turn.
|
||||
#
|
||||
# These are base limits for the FREE tier. Higher tiers (PRO, BUSINESS,
|
||||
# ENTERPRISE) multiply these by their tier multiplier (see
|
||||
# rate_limit.TIER_MULTIPLIERS). User tier is stored in the
|
||||
# User.subscriptionTier DB column and resolved inside
|
||||
# get_global_rate_limits().
|
||||
# TODO: These are deploy-time constants applied identically to every user.
|
||||
# If per-user or per-plan limits are needed (e.g., free tier vs paid), these
|
||||
# must move to the database (e.g., a UserPlan table) and get_usage_status /
|
||||
# check_rate_limit would look up each user's specific limits instead of
|
||||
# reading config.daily_token_limit / config.weekly_token_limit.
|
||||
daily_token_limit: int = Field(
|
||||
default=2_500_000,
|
||||
description="Max tokens per day, resets at midnight UTC (0 = unlimited)",
|
||||
@@ -100,20 +91,6 @@ class ChatConfig(BaseSettings):
|
||||
description="Max tokens per week, resets Monday 00:00 UTC (0 = unlimited)",
|
||||
)
|
||||
|
||||
# Cost (in credits / cents) to reset the daily rate limit using credits.
|
||||
# When a user hits their daily limit, they can spend this amount to reset
|
||||
# the daily counter and keep working. Set to 0 to disable the feature.
|
||||
rate_limit_reset_cost: int = Field(
|
||||
default=500,
|
||||
ge=0,
|
||||
description="Credit cost (in cents) for resetting the daily rate limit. 0 = disabled.",
|
||||
)
|
||||
max_daily_resets: int = Field(
|
||||
default=5,
|
||||
ge=0,
|
||||
description="Maximum number of credit-based rate limit resets per user per day. 0 = unlimited.",
|
||||
)
|
||||
|
||||
# Claude Agent SDK Configuration
|
||||
use_claude_agent_sdk: bool = Field(
|
||||
default=True,
|
||||
@@ -138,32 +115,6 @@ 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. "
|
||||
@@ -213,7 +164,7 @@ class ChatConfig(BaseSettings):
|
||||
|
||||
Single source of truth for "will the SDK route through OpenRouter?".
|
||||
Checks the flag *and* that ``api_key`` + a valid ``base_url`` are
|
||||
present — mirrors the fallback logic in ``build_sdk_env``.
|
||||
present — mirrors the fallback logic in ``_build_sdk_env``.
|
||||
"""
|
||||
if not self.use_openrouter:
|
||||
return False
|
||||
|
||||
@@ -44,32 +44,12 @@ 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
|
||||
# which is retryable. Covers:
|
||||
# - Connection-level: ECONNRESET, dropped TCP connections
|
||||
# - HTTP 429: rate-limit / too-many-requests
|
||||
# - HTTP 5xx: server errors, overloaded
|
||||
# (ECONNRESET / dropped TCP connection) which is retryable.
|
||||
_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"
|
||||
|
||||
@@ -149,8 +149,7 @@ def is_allowed_local_path(path: str, sdk_cwd: str | None = None) -> bool:
|
||||
|
||||
Allowed:
|
||||
- Files under *sdk_cwd* (``/tmp/copilot-<session>/``)
|
||||
- Files under ``~/.claude/projects/<encoded-cwd>/<uuid>/tool-results/...``
|
||||
or ``tool-outputs/...``.
|
||||
- Files under ``~/.claude/projects/<encoded-cwd>/<uuid>/tool-results/...``.
|
||||
The SDK nests tool-results under a conversation UUID directory;
|
||||
the UUID segment is validated with ``_UUID_RE``.
|
||||
"""
|
||||
@@ -175,20 +174,17 @@ def is_allowed_local_path(path: str, sdk_cwd: str | None = None) -> bool:
|
||||
# Defence-in-depth: ensure project_dir didn't escape the base.
|
||||
if not project_dir.startswith(SDK_PROJECTS_DIR + os.sep):
|
||||
return False
|
||||
# Only allow: <encoded-cwd>/<uuid>/<tool-dir>/<file>
|
||||
# Only allow: <encoded-cwd>/<uuid>/tool-results/<file>
|
||||
# The SDK always creates a conversation UUID directory between
|
||||
# the project dir and the tool directory.
|
||||
# Accept both "tool-results" (SDK's persisted outputs) and
|
||||
# "tool-outputs" (the model sometimes confuses workspace paths
|
||||
# with filesystem paths and generates this variant).
|
||||
# the project dir and tool-results/.
|
||||
if resolved.startswith(project_dir + os.sep):
|
||||
relative = resolved[len(project_dir) + 1 :]
|
||||
parts = relative.split(os.sep)
|
||||
# Require exactly: [<uuid>, "tool-results"|"tool-outputs", <file>, ...]
|
||||
# Require exactly: [<uuid>, "tool-results", <file>, ...]
|
||||
if (
|
||||
len(parts) >= 3
|
||||
and _UUID_RE.match(parts[0])
|
||||
and parts[1] in ("tool-results", "tool-outputs")
|
||||
and parts[1] == "tool-results"
|
||||
):
|
||||
return True
|
||||
|
||||
|
||||
@@ -134,21 +134,6 @@ def test_is_allowed_local_path_tool_results_with_uuid():
|
||||
_current_project_dir.set("")
|
||||
|
||||
|
||||
def test_is_allowed_local_path_tool_outputs_with_uuid():
|
||||
"""Files under <encoded-cwd>/<uuid>/tool-outputs/ are also allowed."""
|
||||
encoded = "test-encoded-dir"
|
||||
conv_uuid = "a1b2c3d4-e5f6-7890-abcd-ef1234567890"
|
||||
path = os.path.join(
|
||||
SDK_PROJECTS_DIR, encoded, conv_uuid, "tool-outputs", "output.json"
|
||||
)
|
||||
|
||||
_current_project_dir.set(encoded)
|
||||
try:
|
||||
assert is_allowed_local_path(path, sdk_cwd=None)
|
||||
finally:
|
||||
_current_project_dir.set("")
|
||||
|
||||
|
||||
def test_is_allowed_local_path_tool_results_without_uuid_rejected():
|
||||
"""Direct <encoded-cwd>/tool-results/ (no UUID) is rejected."""
|
||||
encoded = "test-encoded-dir"
|
||||
@@ -174,7 +159,7 @@ def test_is_allowed_local_path_sibling_of_tool_results_is_rejected():
|
||||
|
||||
|
||||
def test_is_allowed_local_path_valid_uuid_wrong_segment_name_rejected():
|
||||
"""A valid UUID dir but non-'tool-results'/'tool-outputs' second segment is rejected."""
|
||||
"""A valid UUID dir but non-'tool-results' second segment is rejected."""
|
||||
encoded = "test-encoded-dir"
|
||||
uuid_str = "12345678-1234-5678-9abc-def012345678"
|
||||
path = os.path.join(
|
||||
|
||||
@@ -18,13 +18,7 @@ from prisma.types import (
|
||||
from backend.data import db
|
||||
from backend.util.json import SafeJson, sanitize_string
|
||||
|
||||
from .model import (
|
||||
ChatMessage,
|
||||
ChatSession,
|
||||
ChatSessionInfo,
|
||||
ChatSessionMetadata,
|
||||
invalidate_session_cache,
|
||||
)
|
||||
from .model import ChatMessage, ChatSession, ChatSessionInfo
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@@ -41,7 +35,6 @@ async def get_chat_session(session_id: str) -> ChatSession | None:
|
||||
async def create_chat_session(
|
||||
session_id: str,
|
||||
user_id: str,
|
||||
metadata: ChatSessionMetadata | None = None,
|
||||
) -> ChatSessionInfo:
|
||||
"""Create a new chat session in the database."""
|
||||
data = ChatSessionCreateInput(
|
||||
@@ -50,7 +43,6 @@ async def create_chat_session(
|
||||
credentials=SafeJson({}),
|
||||
successfulAgentRuns=SafeJson({}),
|
||||
successfulAgentSchedules=SafeJson({}),
|
||||
metadata=SafeJson((metadata or ChatSessionMetadata()).model_dump()),
|
||||
)
|
||||
prisma_session = await PrismaChatSession.prisma().create(data=data)
|
||||
return ChatSessionInfo.from_db(prisma_session)
|
||||
@@ -65,12 +57,7 @@ async def update_chat_session(
|
||||
total_completion_tokens: int | None = None,
|
||||
title: str | None = None,
|
||||
) -> ChatSession | None:
|
||||
"""Update a chat session's mutable fields.
|
||||
|
||||
Note: ``metadata`` (which includes ``dry_run``) is intentionally omitted —
|
||||
it is set once at creation time and treated as immutable for the lifetime
|
||||
of the session.
|
||||
"""
|
||||
"""Update a chat session's metadata."""
|
||||
data: ChatSessionUpdateInput = {"updatedAt": datetime.now(UTC)}
|
||||
|
||||
if credentials is not None:
|
||||
@@ -230,9 +217,6 @@ async def add_chat_messages_batch(
|
||||
if msg.get("function_call") is not None:
|
||||
data["functionCall"] = SafeJson(msg["function_call"])
|
||||
|
||||
if msg.get("duration_ms") is not None:
|
||||
data["durationMs"] = msg["duration_ms"]
|
||||
|
||||
messages_data.append(data)
|
||||
|
||||
# Run create_many and session update in parallel within transaction
|
||||
@@ -375,22 +359,3 @@ async def update_tool_message_content(
|
||||
f"tool_call_id {tool_call_id}: {e}"
|
||||
)
|
||||
return False
|
||||
|
||||
|
||||
async def set_turn_duration(session_id: str, duration_ms: int) -> None:
|
||||
"""Set durationMs on the last assistant message in a session.
|
||||
|
||||
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"},
|
||||
order={"sequence": "desc"},
|
||||
)
|
||||
if last_msg:
|
||||
await PrismaChatMessage.prisma().update(
|
||||
where={"id": last_msg.id},
|
||||
data={"durationMs": duration_ms},
|
||||
)
|
||||
# Invalidate cache so the session is re-fetched from DB with durationMs
|
||||
await invalidate_session_cache(session_id)
|
||||
|
||||
@@ -251,31 +251,20 @@ class CoPilotProcessor:
|
||||
stream_fn = stream_chat_completion_dummy
|
||||
log.warning("Using DUMMY service (CHAT_TEST_MODE=true)")
|
||||
else:
|
||||
# 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,
|
||||
)
|
||||
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
|
||||
else stream_chat_completion_baseline
|
||||
)
|
||||
log.info(
|
||||
f"Using {'SDK' if use_sdk else 'baseline'} service "
|
||||
f"(mode={entry.mode or 'default'})"
|
||||
)
|
||||
log.info(f"Using {'SDK' if use_sdk else 'baseline'} service")
|
||||
|
||||
# Stream chat completion and publish chunks to Redis.
|
||||
# stream_and_publish wraps the raw stream with registry
|
||||
|
||||
@@ -6,7 +6,6 @@ Defines two exchanges and queues following the graph executor pattern:
|
||||
"""
|
||||
|
||||
import logging
|
||||
from typing import Literal
|
||||
|
||||
from pydantic import BaseModel
|
||||
|
||||
@@ -157,9 +156,6 @@ class CoPilotExecutionEntry(BaseModel):
|
||||
file_ids: list[str] | None = None
|
||||
"""Workspace file IDs attached to the user's message"""
|
||||
|
||||
mode: Literal["fast", "extended_thinking"] | None = None
|
||||
"""Autopilot mode override: 'fast' or 'extended_thinking'. None = server default."""
|
||||
|
||||
|
||||
class CancelCoPilotEvent(BaseModel):
|
||||
"""Event to cancel a CoPilot operation."""
|
||||
@@ -179,7 +175,6 @@ async def enqueue_copilot_turn(
|
||||
is_user_message: bool = True,
|
||||
context: dict[str, str] | None = None,
|
||||
file_ids: list[str] | None = None,
|
||||
mode: Literal["fast", "extended_thinking"] | None = None,
|
||||
) -> None:
|
||||
"""Enqueue a CoPilot task for processing by the executor service.
|
||||
|
||||
@@ -191,7 +186,6 @@ async def enqueue_copilot_turn(
|
||||
is_user_message: Whether the message is from the user (vs system/assistant)
|
||||
context: Optional context for the message (e.g., {url: str, content: str})
|
||||
file_ids: Optional workspace file IDs attached to the user's message
|
||||
mode: Autopilot mode override ('fast' or 'extended_thinking'). None = server default.
|
||||
"""
|
||||
from backend.util.clients import get_async_copilot_queue
|
||||
|
||||
@@ -203,7 +197,6 @@ async def enqueue_copilot_turn(
|
||||
is_user_message=is_user_message,
|
||||
context=context,
|
||||
file_ids=file_ids,
|
||||
mode=mode,
|
||||
)
|
||||
|
||||
queue_client = await get_async_copilot_queue()
|
||||
|
||||
@@ -59,16 +59,6 @@ _null_cache: TTLCache[tuple[str, str], bool] = TTLCache(
|
||||
maxsize=_CACHE_MAX_SIZE, ttl=_NULL_CACHE_TTL
|
||||
)
|
||||
|
||||
# GitHub user identity caches (keyed by user_id only, not provider tuple).
|
||||
# Declared here so invalidate_user_provider_cache() can reference them.
|
||||
_GH_IDENTITY_CACHE_TTL = 600.0 # 10 min — profile data rarely changes
|
||||
_gh_identity_cache: TTLCache[str, dict[str, str]] = TTLCache(
|
||||
maxsize=_CACHE_MAX_SIZE, ttl=_GH_IDENTITY_CACHE_TTL
|
||||
)
|
||||
_gh_identity_null_cache: TTLCache[str, bool] = TTLCache(
|
||||
maxsize=_CACHE_MAX_SIZE, ttl=_NULL_CACHE_TTL
|
||||
)
|
||||
|
||||
|
||||
def invalidate_user_provider_cache(user_id: str, provider: str) -> None:
|
||||
"""Remove the cached entry for *user_id*/*provider* from both caches.
|
||||
@@ -76,19 +66,11 @@ def invalidate_user_provider_cache(user_id: str, provider: str) -> None:
|
||||
Call this after storing new credentials so that the next
|
||||
``get_provider_token()`` call performs a fresh DB lookup instead of
|
||||
serving a stale TTL-cached result.
|
||||
|
||||
For GitHub specifically, also clears the git-identity caches so that
|
||||
``get_github_user_git_identity()`` re-fetches the user's profile on
|
||||
the next call instead of serving stale identity data.
|
||||
"""
|
||||
key = (user_id, provider)
|
||||
_token_cache.pop(key, None)
|
||||
_null_cache.pop(key, None)
|
||||
|
||||
if provider == "github":
|
||||
_gh_identity_cache.pop(user_id, None)
|
||||
_gh_identity_null_cache.pop(user_id, None)
|
||||
|
||||
|
||||
# Register this module's cache-bust function with the credentials manager so
|
||||
# that any create/update/delete operation immediately evicts stale cache
|
||||
@@ -141,7 +123,6 @@ async def get_provider_token(user_id: str, provider: str) -> str | None:
|
||||
[c for c in creds_list if c.type == "oauth2"],
|
||||
key=lambda c: 0 if "repo" in (cast(OAuth2Credentials, c).scopes or []) else 1,
|
||||
)
|
||||
refresh_failed = False
|
||||
for creds in oauth2_creds:
|
||||
if creds.type == "oauth2":
|
||||
try:
|
||||
@@ -160,7 +141,6 @@ async def get_provider_token(user_id: str, provider: str) -> str | None:
|
||||
# Do NOT fall back to the stale token — it is likely expired
|
||||
# or revoked. Returning None forces the caller to re-auth,
|
||||
# preventing the LLM from receiving a non-functional token.
|
||||
refresh_failed = True
|
||||
continue
|
||||
_token_cache[cache_key] = token
|
||||
return token
|
||||
@@ -172,12 +152,8 @@ async def get_provider_token(user_id: str, provider: str) -> str | None:
|
||||
_token_cache[cache_key] = token
|
||||
return token
|
||||
|
||||
# Only cache "not connected" when the user truly has no credentials for this
|
||||
# provider. If we had OAuth credentials but refresh failed (e.g. transient
|
||||
# network error, event-loop mismatch), do NOT cache the negative result —
|
||||
# the next call should retry the refresh instead of being blocked for 60 s.
|
||||
if not refresh_failed:
|
||||
_null_cache[cache_key] = True
|
||||
# No credentials found — cache to avoid repeated DB hits.
|
||||
_null_cache[cache_key] = True
|
||||
return None
|
||||
|
||||
|
||||
@@ -195,76 +171,3 @@ async def get_integration_env_vars(user_id: str) -> dict[str, str]:
|
||||
for var in var_names:
|
||||
env[var] = token
|
||||
return env
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# GitHub user identity (for git committer env vars)
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
async def get_github_user_git_identity(user_id: str) -> dict[str, str] | None:
|
||||
"""Fetch the GitHub user's name and email for git committer env vars.
|
||||
|
||||
Uses the ``/user`` GitHub API endpoint with the user's stored token.
|
||||
Returns a dict with ``GIT_AUTHOR_NAME``, ``GIT_AUTHOR_EMAIL``,
|
||||
``GIT_COMMITTER_NAME``, and ``GIT_COMMITTER_EMAIL`` if the user has a
|
||||
connected GitHub account. Returns ``None`` otherwise.
|
||||
|
||||
Results are cached for 10 minutes; "not connected" results are cached for
|
||||
60 s (same as null-token cache).
|
||||
"""
|
||||
if user_id in _gh_identity_null_cache:
|
||||
return None
|
||||
if cached := _gh_identity_cache.get(user_id):
|
||||
return cached
|
||||
|
||||
token = await get_provider_token(user_id, "github")
|
||||
if not token:
|
||||
_gh_identity_null_cache[user_id] = True
|
||||
return None
|
||||
|
||||
import aiohttp
|
||||
|
||||
try:
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.get(
|
||||
"https://api.github.com/user",
|
||||
headers={
|
||||
"Authorization": f"token {token}",
|
||||
"Accept": "application/vnd.github+json",
|
||||
},
|
||||
timeout=aiohttp.ClientTimeout(total=5),
|
||||
) as resp:
|
||||
if resp.status != 200:
|
||||
logger.warning(
|
||||
"[git-identity] GitHub /user returned %s for user %s",
|
||||
resp.status,
|
||||
user_id,
|
||||
)
|
||||
return None
|
||||
data = await resp.json()
|
||||
except Exception as exc:
|
||||
logger.warning(
|
||||
"[git-identity] Failed to fetch GitHub profile for user %s: %s",
|
||||
user_id,
|
||||
exc,
|
||||
)
|
||||
return None
|
||||
|
||||
name = data.get("name") or data.get("login") or "AutoGPT User"
|
||||
# GitHub may return email=null if the user has set their email to private.
|
||||
# Fall back to the noreply address GitHub generates for every account.
|
||||
email = data.get("email")
|
||||
if not email:
|
||||
gh_id = data.get("id", "")
|
||||
login = data.get("login", "user")
|
||||
email = f"{gh_id}+{login}@users.noreply.github.com"
|
||||
|
||||
identity = {
|
||||
"GIT_AUTHOR_NAME": name,
|
||||
"GIT_AUTHOR_EMAIL": email,
|
||||
"GIT_COMMITTER_NAME": name,
|
||||
"GIT_COMMITTER_EMAIL": email,
|
||||
}
|
||||
_gh_identity_cache[user_id] = identity
|
||||
return identity
|
||||
|
||||
@@ -9,8 +9,6 @@ from backend.copilot.integration_creds import (
|
||||
_NULL_CACHE_TTL,
|
||||
_TOKEN_CACHE_TTL,
|
||||
PROVIDER_ENV_VARS,
|
||||
_gh_identity_cache,
|
||||
_gh_identity_null_cache,
|
||||
_null_cache,
|
||||
_token_cache,
|
||||
get_integration_env_vars,
|
||||
@@ -51,13 +49,9 @@ def clear_caches():
|
||||
"""Ensure clean caches before and after every test."""
|
||||
_token_cache.clear()
|
||||
_null_cache.clear()
|
||||
_gh_identity_cache.clear()
|
||||
_gh_identity_null_cache.clear()
|
||||
yield
|
||||
_token_cache.clear()
|
||||
_null_cache.clear()
|
||||
_gh_identity_cache.clear()
|
||||
_gh_identity_null_cache.clear()
|
||||
|
||||
|
||||
class TestInvalidateUserProviderCache:
|
||||
@@ -83,34 +77,6 @@ class TestInvalidateUserProviderCache:
|
||||
invalidate_user_provider_cache(_USER, _PROVIDER)
|
||||
assert other_key in _token_cache
|
||||
|
||||
def test_clears_gh_identity_cache_for_github_provider(self):
|
||||
"""When provider is 'github', identity caches must also be cleared."""
|
||||
_gh_identity_cache[_USER] = {
|
||||
"GIT_AUTHOR_NAME": "Old Name",
|
||||
"GIT_AUTHOR_EMAIL": "old@example.com",
|
||||
"GIT_COMMITTER_NAME": "Old Name",
|
||||
"GIT_COMMITTER_EMAIL": "old@example.com",
|
||||
}
|
||||
invalidate_user_provider_cache(_USER, "github")
|
||||
assert _USER not in _gh_identity_cache
|
||||
|
||||
def test_clears_gh_identity_null_cache_for_github_provider(self):
|
||||
"""When provider is 'github', the identity null-cache must also be cleared."""
|
||||
_gh_identity_null_cache[_USER] = True
|
||||
invalidate_user_provider_cache(_USER, "github")
|
||||
assert _USER not in _gh_identity_null_cache
|
||||
|
||||
def test_does_not_clear_gh_identity_cache_for_other_providers(self):
|
||||
"""When provider is NOT 'github', identity caches must be left alone."""
|
||||
_gh_identity_cache[_USER] = {
|
||||
"GIT_AUTHOR_NAME": "Some Name",
|
||||
"GIT_AUTHOR_EMAIL": "some@example.com",
|
||||
"GIT_COMMITTER_NAME": "Some Name",
|
||||
"GIT_COMMITTER_EMAIL": "some@example.com",
|
||||
}
|
||||
invalidate_user_provider_cache(_USER, "some-other-provider")
|
||||
assert _USER in _gh_identity_cache
|
||||
|
||||
|
||||
class TestGetProviderToken:
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
@@ -163,15 +129,8 @@ class TestGetProviderToken:
|
||||
assert result == "oauth-tok"
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_oauth2_refresh_failure_returns_none_without_null_cache(self):
|
||||
"""On refresh failure, return None but do NOT cache in null_cache.
|
||||
|
||||
The user has credentials — they just couldn't be refreshed right now
|
||||
(e.g. transient network error or event-loop mismatch in the copilot
|
||||
executor). Caching a negative result would block all credential
|
||||
lookups for 60 s even though the creds exist and may refresh fine
|
||||
on the next attempt.
|
||||
"""
|
||||
async def test_oauth2_refresh_failure_returns_none(self):
|
||||
"""On refresh failure, return None instead of caching a stale token."""
|
||||
oauth_creds = _make_oauth2_creds("stale-oauth-tok")
|
||||
mock_manager = MagicMock()
|
||||
mock_manager.store.get_creds_by_provider = AsyncMock(return_value=[oauth_creds])
|
||||
@@ -182,8 +141,6 @@ class TestGetProviderToken:
|
||||
|
||||
# Stale tokens must NOT be returned — forces re-auth.
|
||||
assert result is None
|
||||
# Must NOT cache negative result when refresh failed — next call retries.
|
||||
assert (_USER, _PROVIDER) not in _null_cache
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_no_credentials_caches_null_entry(self):
|
||||
@@ -219,96 +176,6 @@ class TestGetProviderToken:
|
||||
assert _NULL_CACHE_TTL < _TOKEN_CACHE_TTL
|
||||
|
||||
|
||||
class TestThreadSafetyLocks:
|
||||
"""Bug reproduction: shared AsyncRedisKeyedMutex across threads caused
|
||||
'Future attached to a different loop' when copilot workers accessed
|
||||
credentials from different event loops."""
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_store_locks_returns_per_thread_instance(self):
|
||||
"""IntegrationCredentialsStore.locks() must return different instances
|
||||
for different threads (via @thread_cached)."""
|
||||
import asyncio
|
||||
import concurrent.futures
|
||||
|
||||
from backend.integrations.credentials_store import IntegrationCredentialsStore
|
||||
|
||||
store = IntegrationCredentialsStore()
|
||||
|
||||
async def get_locks_id():
|
||||
mock_redis = AsyncMock()
|
||||
with patch(
|
||||
"backend.integrations.credentials_store.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
locks = await store.locks()
|
||||
return id(locks)
|
||||
|
||||
# Get locks from main thread
|
||||
main_id = await get_locks_id()
|
||||
|
||||
# Get locks from a worker thread
|
||||
def run_in_thread():
|
||||
loop = asyncio.new_event_loop()
|
||||
try:
|
||||
return loop.run_until_complete(get_locks_id())
|
||||
finally:
|
||||
loop.close()
|
||||
|
||||
with concurrent.futures.ThreadPoolExecutor(max_workers=1) as pool:
|
||||
worker_id = await asyncio.get_event_loop().run_in_executor(
|
||||
pool, run_in_thread
|
||||
)
|
||||
|
||||
assert main_id != worker_id, (
|
||||
"Store.locks() returned the same instance across threads. "
|
||||
"This would cause 'Future attached to a different loop' errors."
|
||||
)
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_manager_delegates_to_store_locks(self):
|
||||
"""IntegrationCredentialsManager.locks() should delegate to store."""
|
||||
from backend.integrations.creds_manager import IntegrationCredentialsManager
|
||||
|
||||
manager = IntegrationCredentialsManager()
|
||||
mock_redis = AsyncMock()
|
||||
|
||||
with patch(
|
||||
"backend.integrations.credentials_store.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
locks = await manager.locks()
|
||||
|
||||
# Should have gotten it from the store
|
||||
assert locks is not None
|
||||
|
||||
|
||||
class TestRefreshUnlockedPath:
|
||||
"""Bug reproduction: copilot worker threads need lock-free refresh because
|
||||
Redis-backed asyncio.Lock created on one event loop can't be used on another."""
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_refresh_if_needed_lock_false_skips_redis(self):
|
||||
"""refresh_if_needed(lock=False) must not touch Redis locks at all."""
|
||||
from backend.integrations.creds_manager import IntegrationCredentialsManager
|
||||
|
||||
manager = IntegrationCredentialsManager()
|
||||
creds = _make_oauth2_creds()
|
||||
|
||||
mock_handler = MagicMock()
|
||||
mock_handler.needs_refresh = MagicMock(return_value=False)
|
||||
|
||||
with patch(
|
||||
"backend.integrations.creds_manager._get_provider_oauth_handler",
|
||||
new_callable=AsyncMock,
|
||||
return_value=mock_handler,
|
||||
):
|
||||
result = await manager.refresh_if_needed(_USER, creds, lock=False)
|
||||
|
||||
# Should return credentials without touching locks
|
||||
assert result.id == creds.id
|
||||
|
||||
|
||||
class TestGetIntegrationEnvVars:
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_injects_all_env_vars_for_provider(self):
|
||||
|
||||
@@ -46,16 +46,6 @@ def _get_session_cache_key(session_id: str) -> str:
|
||||
# ===================== Chat data models ===================== #
|
||||
|
||||
|
||||
class ChatSessionMetadata(BaseModel):
|
||||
"""Typed metadata stored in the ``metadata`` JSON column of ChatSession.
|
||||
|
||||
Add new session-level flags here instead of adding DB columns —
|
||||
no migration required for new fields as long as a default is provided.
|
||||
"""
|
||||
|
||||
dry_run: bool = False
|
||||
|
||||
|
||||
class ChatMessage(BaseModel):
|
||||
role: str
|
||||
content: str | None = None
|
||||
@@ -64,7 +54,6 @@ class ChatMessage(BaseModel):
|
||||
refusal: str | None = None
|
||||
tool_calls: list[dict] | None = None
|
||||
function_call: dict | None = None
|
||||
duration_ms: int | None = None
|
||||
|
||||
@staticmethod
|
||||
def from_db(prisma_message: PrismaChatMessage) -> "ChatMessage":
|
||||
@@ -77,7 +66,6 @@ class ChatMessage(BaseModel):
|
||||
refusal=prisma_message.refusal,
|
||||
tool_calls=_parse_json_field(prisma_message.toolCalls),
|
||||
function_call=_parse_json_field(prisma_message.functionCall),
|
||||
duration_ms=prisma_message.durationMs,
|
||||
)
|
||||
|
||||
|
||||
@@ -100,12 +88,6 @@ class ChatSessionInfo(BaseModel):
|
||||
updated_at: datetime
|
||||
successful_agent_runs: dict[str, int] = {}
|
||||
successful_agent_schedules: dict[str, int] = {}
|
||||
metadata: ChatSessionMetadata = ChatSessionMetadata()
|
||||
|
||||
@property
|
||||
def dry_run(self) -> bool:
|
||||
"""Convenience accessor for ``metadata.dry_run``."""
|
||||
return self.metadata.dry_run
|
||||
|
||||
@classmethod
|
||||
def from_db(cls, prisma_session: PrismaChatSession) -> Self:
|
||||
@@ -119,10 +101,6 @@ class ChatSessionInfo(BaseModel):
|
||||
prisma_session.successfulAgentSchedules, default={}
|
||||
)
|
||||
|
||||
# Parse typed metadata from the JSON column.
|
||||
raw_metadata = _parse_json_field(prisma_session.metadata, default={})
|
||||
metadata = ChatSessionMetadata.model_validate(raw_metadata)
|
||||
|
||||
# Calculate usage from token counts.
|
||||
# NOTE: Per-turn cache_read_tokens / cache_creation_tokens breakdown
|
||||
# is lost after persistence — the DB only stores aggregate prompt and
|
||||
@@ -148,7 +126,6 @@ class ChatSessionInfo(BaseModel):
|
||||
updated_at=prisma_session.updatedAt,
|
||||
successful_agent_runs=successful_agent_runs,
|
||||
successful_agent_schedules=successful_agent_schedules,
|
||||
metadata=metadata,
|
||||
)
|
||||
|
||||
|
||||
@@ -156,7 +133,7 @@ class ChatSession(ChatSessionInfo):
|
||||
messages: list[ChatMessage]
|
||||
|
||||
@classmethod
|
||||
def new(cls, user_id: str, *, dry_run: bool) -> Self:
|
||||
def new(cls, user_id: str) -> Self:
|
||||
return cls(
|
||||
session_id=str(uuid.uuid4()),
|
||||
user_id=user_id,
|
||||
@@ -166,7 +143,6 @@ class ChatSession(ChatSessionInfo):
|
||||
credentials={},
|
||||
started_at=datetime.now(UTC),
|
||||
updated_at=datetime.now(UTC),
|
||||
metadata=ChatSessionMetadata(dry_run=dry_run),
|
||||
)
|
||||
|
||||
@classmethod
|
||||
@@ -554,7 +530,6 @@ async def _save_session_to_db(
|
||||
await db.create_chat_session(
|
||||
session_id=session.session_id,
|
||||
user_id=session.user_id,
|
||||
metadata=session.metadata,
|
||||
)
|
||||
existing_message_count = 0
|
||||
|
||||
@@ -632,27 +607,21 @@ async def append_and_save_message(session_id: str, message: ChatMessage) -> Chat
|
||||
return session
|
||||
|
||||
|
||||
async def create_chat_session(user_id: str, *, dry_run: bool) -> ChatSession:
|
||||
async def create_chat_session(user_id: str) -> ChatSession:
|
||||
"""Create a new chat session and persist it.
|
||||
|
||||
Args:
|
||||
user_id: The authenticated user ID.
|
||||
dry_run: When True, run_block and run_agent tool calls in this
|
||||
session are forced to use dry-run simulation mode.
|
||||
|
||||
Raises:
|
||||
DatabaseError: If the database write fails. We fail fast to ensure
|
||||
callers never receive a non-persisted session that only exists
|
||||
in cache (which would be lost when the cache expires).
|
||||
"""
|
||||
session = ChatSession.new(user_id, dry_run=dry_run)
|
||||
session = ChatSession.new(user_id)
|
||||
|
||||
# Create in database first - fail fast if this fails
|
||||
try:
|
||||
await chat_db().create_chat_session(
|
||||
session_id=session.session_id,
|
||||
user_id=user_id,
|
||||
metadata=session.metadata,
|
||||
)
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to create session {session.session_id} in database: {e}")
|
||||
|
||||
@@ -46,7 +46,7 @@ messages = [
|
||||
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_chatsession_serialization_deserialization():
|
||||
s = ChatSession.new(user_id="abc123", dry_run=False)
|
||||
s = ChatSession.new(user_id="abc123")
|
||||
s.messages = messages
|
||||
s.usage = [Usage(prompt_tokens=100, completion_tokens=200, total_tokens=300)]
|
||||
serialized = s.model_dump_json()
|
||||
@@ -57,7 +57,7 @@ async def test_chatsession_serialization_deserialization():
|
||||
@pytest.mark.asyncio(loop_scope="session")
|
||||
async def test_chatsession_redis_storage(setup_test_user, test_user_id):
|
||||
|
||||
s = ChatSession.new(user_id=test_user_id, dry_run=False)
|
||||
s = ChatSession.new(user_id=test_user_id)
|
||||
s.messages = messages
|
||||
|
||||
s = await upsert_chat_session(s)
|
||||
@@ -75,7 +75,7 @@ async def test_chatsession_redis_storage_user_id_mismatch(
|
||||
setup_test_user, test_user_id
|
||||
):
|
||||
|
||||
s = ChatSession.new(user_id=test_user_id, dry_run=False)
|
||||
s = ChatSession.new(user_id=test_user_id)
|
||||
s.messages = messages
|
||||
s = await upsert_chat_session(s)
|
||||
|
||||
@@ -90,7 +90,7 @@ async def test_chatsession_db_storage(setup_test_user, test_user_id):
|
||||
from backend.data.redis_client import get_redis_async
|
||||
|
||||
# Create session with messages including assistant message
|
||||
s = ChatSession.new(user_id=test_user_id, dry_run=False)
|
||||
s = ChatSession.new(user_id=test_user_id)
|
||||
s.messages = messages # Contains user, assistant, and tool messages
|
||||
assert s.session_id is not None, "Session id is not set"
|
||||
# Upsert to save to both cache and DB
|
||||
@@ -241,7 +241,7 @@ _raw_tc2 = {
|
||||
|
||||
def test_add_tool_call_appends_to_existing_assistant():
|
||||
"""When the last assistant is from the current turn, tool_call is added to it."""
|
||||
session = ChatSession.new(user_id="u", dry_run=False)
|
||||
session = ChatSession.new(user_id="u")
|
||||
session.messages = [
|
||||
ChatMessage(role="user", content="hi"),
|
||||
ChatMessage(role="assistant", content="working on it"),
|
||||
@@ -254,7 +254,7 @@ def test_add_tool_call_appends_to_existing_assistant():
|
||||
|
||||
def test_add_tool_call_creates_assistant_when_none_exists():
|
||||
"""When there's no current-turn assistant, a new one is created."""
|
||||
session = ChatSession.new(user_id="u", dry_run=False)
|
||||
session = ChatSession.new(user_id="u")
|
||||
session.messages = [
|
||||
ChatMessage(role="user", content="hi"),
|
||||
]
|
||||
@@ -267,7 +267,7 @@ def test_add_tool_call_creates_assistant_when_none_exists():
|
||||
|
||||
def test_add_tool_call_does_not_cross_user_boundary():
|
||||
"""A user message acts as a boundary — previous assistant is not modified."""
|
||||
session = ChatSession.new(user_id="u", dry_run=False)
|
||||
session = ChatSession.new(user_id="u")
|
||||
session.messages = [
|
||||
ChatMessage(role="assistant", content="old turn"),
|
||||
ChatMessage(role="user", content="new message"),
|
||||
@@ -282,7 +282,7 @@ def test_add_tool_call_does_not_cross_user_boundary():
|
||||
|
||||
def test_add_tool_call_multiple_times():
|
||||
"""Multiple long-running tool calls accumulate on the same assistant."""
|
||||
session = ChatSession.new(user_id="u", dry_run=False)
|
||||
session = ChatSession.new(user_id="u")
|
||||
session.messages = [
|
||||
ChatMessage(role="user", content="hi"),
|
||||
ChatMessage(role="assistant", content="doing stuff"),
|
||||
@@ -300,7 +300,7 @@ def test_add_tool_call_multiple_times():
|
||||
|
||||
def test_to_openai_messages_merges_split_assistants():
|
||||
"""End-to-end: session with split assistants produces valid OpenAI messages."""
|
||||
session = ChatSession.new(user_id="u", dry_run=False)
|
||||
session = ChatSession.new(user_id="u")
|
||||
session.messages = [
|
||||
ChatMessage(role="user", content="build agent"),
|
||||
ChatMessage(role="assistant", content="Let me build that"),
|
||||
@@ -352,7 +352,7 @@ async def test_concurrent_saves_collision_detection(setup_test_user, test_user_i
|
||||
import asyncio
|
||||
|
||||
# Create a session with initial messages
|
||||
session = ChatSession.new(user_id=test_user_id, dry_run=False)
|
||||
session = ChatSession.new(user_id=test_user_id)
|
||||
for i in range(3):
|
||||
session.messages.append(
|
||||
ChatMessage(
|
||||
|
||||
@@ -66,7 +66,6 @@ from pydantic import BaseModel, PrivateAttr
|
||||
ToolName = Literal[
|
||||
# Platform tools (must match keys in TOOL_REGISTRY)
|
||||
"add_understanding",
|
||||
"ask_question",
|
||||
"bash_exec",
|
||||
"browser_act",
|
||||
"browser_navigate",
|
||||
@@ -103,7 +102,6 @@ ToolName = Literal[
|
||||
"web_fetch",
|
||||
"write_workspace_file",
|
||||
# SDK built-ins
|
||||
"Agent",
|
||||
"Edit",
|
||||
"Glob",
|
||||
"Grep",
|
||||
|
||||
@@ -544,7 +544,6 @@ class TestApplyToolPermissions:
|
||||
class TestSdkBuiltinToolNames:
|
||||
def test_expected_builtins_present(self):
|
||||
expected = {
|
||||
"Agent",
|
||||
"Read",
|
||||
"Write",
|
||||
"Edit",
|
||||
|
||||
@@ -18,18 +18,6 @@ After `write_workspace_file`, embed the `download_url` in Markdown:
|
||||
- Image: ``
|
||||
- Video: ``
|
||||
|
||||
### Handling binary/image data in tool outputs — CRITICAL
|
||||
When a tool output contains base64-encoded binary data (images, PDFs, etc.):
|
||||
1. **NEVER** try to inline or render the base64 content in your response.
|
||||
2. **Save** the data to workspace using `write_workspace_file` (pass the base64 data URI as content).
|
||||
3. **Show** the result via the workspace download URL in Markdown: ``.
|
||||
|
||||
### Passing large data between tools — CRITICAL
|
||||
When tool outputs produce large text that you need to feed into another tool:
|
||||
- **NEVER** copy-paste the full text into the next tool call argument.
|
||||
- **Save** the output to a file (workspace or local), then use `@@agptfile:` references.
|
||||
- This avoids token limits and ensures data integrity.
|
||||
|
||||
### File references — @@agptfile:
|
||||
Pass large file content to tools by reference: `@@agptfile:<uri>[<start>-<end>]`
|
||||
- `workspace://<file_id>` or `workspace:///<path>` — workspace files
|
||||
@@ -119,28 +107,6 @@ Do not re-fetch or re-generate data you already have from prior tool calls.
|
||||
After building the file, reference it with `@@agptfile:` in other tools:
|
||||
`@@agptfile:/home/user/report.md`
|
||||
|
||||
### Web search best practices
|
||||
- If 3 similar web searches don't return the specific data you need, conclude
|
||||
it isn't publicly available and work with what you have.
|
||||
- Prefer fewer, well-targeted searches over many variations of the same query.
|
||||
- When spawning sub-agents for research, ensure each has a distinct
|
||||
non-overlapping scope to avoid redundant searches.
|
||||
|
||||
|
||||
### Tool Discovery Priority
|
||||
|
||||
When the user asks to interact with a service or API, follow this order:
|
||||
|
||||
1. **find_block first** — Search platform blocks with `find_block`. The platform has hundreds of built-in blocks (Google Sheets, Docs, Calendar, Gmail, Slack, GitHub, etc.) that work without extra setup.
|
||||
|
||||
2. **run_mcp_tool** — If no matching block exists, check if a hosted MCP server is available for the service. Only use known MCP server URLs from the registry.
|
||||
|
||||
3. **SendAuthenticatedWebRequestBlock** — If no block or MCP server exists, use `SendAuthenticatedWebRequestBlock` with existing host-scoped credentials. Check available credentials via `connect_integration`.
|
||||
|
||||
4. **Manual API call** — As a last resort, guide the user to set up credentials and use `SendAuthenticatedWebRequestBlock` with direct API calls.
|
||||
|
||||
**Never skip step 1.** Built-in blocks are more reliable, tested, and user-friendly than MCP or raw API calls.
|
||||
|
||||
### Sub-agent tasks
|
||||
- When using the Task tool, NEVER set `run_in_background` to true.
|
||||
All tasks must run in the foreground.
|
||||
@@ -165,11 +131,6 @@ parent autopilot handles orchestration.
|
||||
# E2B-only notes — E2B has full internet access so gh CLI works there.
|
||||
# Not shown in local (bubblewrap) mode: --unshare-net blocks all network.
|
||||
_E2B_TOOL_NOTES = """
|
||||
### SDK tool-result files in E2B
|
||||
When you `Read` an SDK tool-result file, it is automatically copied into the
|
||||
sandbox so `bash_exec` can access it for further processing.
|
||||
The exact sandbox path is shown in the `[Sandbox copy available at ...]` note.
|
||||
|
||||
### GitHub CLI (`gh`) and git
|
||||
- If the user has connected their GitHub account, both `gh` and `git` are
|
||||
pre-authenticated — use them directly without any manual login step.
|
||||
@@ -235,22 +196,18 @@ def _build_storage_supplement(
|
||||
- Files here **survive across sessions indefinitely**
|
||||
|
||||
### Moving files between storages
|
||||
- **{file_move_name_1_to_2}**: `write_workspace_file(filename="output.json", source_path="/path/to/local/file")`
|
||||
- **{file_move_name_2_to_1}**: `read_workspace_file(path="tool-outputs/data.json", save_to_path="{working_dir}/data.json")`
|
||||
- **{file_move_name_1_to_2}**: Copy to persistent workspace
|
||||
- **{file_move_name_2_to_1}**: Download for processing
|
||||
|
||||
### File persistence
|
||||
Important files (code, configs, outputs) should be saved to workspace to ensure they persist.
|
||||
|
||||
### SDK tool-result files
|
||||
When tool outputs are large, the SDK truncates them and saves the full output to
|
||||
a local file under `~/.claude/projects/.../tool-results/` (or `tool-outputs/`).
|
||||
To read these files, use `Read` — it reads from the host filesystem.
|
||||
|
||||
### Large tool outputs saved to workspace
|
||||
When a tool output contains `<tool-output-truncated workspace_path="...">`, the
|
||||
full output is in workspace storage (NOT on the local filesystem). To access it:
|
||||
- Use `read_workspace_file(path="...", offset=..., length=50000)` for reading sections.
|
||||
- To process in the sandbox, use `read_workspace_file(path="...", save_to_path="{working_dir}/file.json")` first, then use `bash_exec` on the local copy.
|
||||
a local file under `~/.claude/projects/.../tool-results/`. To read these files,
|
||||
always use `read_file` or `Read` (NOT `read_workspace_file`).
|
||||
`read_workspace_file` reads from cloud workspace storage, where SDK
|
||||
tool-results are NOT stored.
|
||||
{_SHARED_TOOL_NOTES}{extra_notes}"""
|
||||
|
||||
|
||||
|
||||
@@ -1,28 +0,0 @@
|
||||
"""Tests for agent generation guide — verifies clarification section."""
|
||||
|
||||
from pathlib import Path
|
||||
|
||||
|
||||
class TestAgentGenerationGuideContainsClarifySection:
|
||||
"""The agent generation guide must include the clarification section."""
|
||||
|
||||
def test_guide_includes_clarify_section(self):
|
||||
guide_path = Path(__file__).parent / "sdk" / "agent_generation_guide.md"
|
||||
content = guide_path.read_text(encoding="utf-8")
|
||||
assert "Before or During Building" in content
|
||||
|
||||
def test_guide_mentions_find_block_for_clarification(self):
|
||||
guide_path = Path(__file__).parent / "sdk" / "agent_generation_guide.md"
|
||||
content = guide_path.read_text(encoding="utf-8")
|
||||
clarify_section = content.split("Before or During Building")[1].split(
|
||||
"### Workflow"
|
||||
)[0]
|
||||
assert "find_block" in clarify_section
|
||||
|
||||
def test_guide_mentions_ask_question_tool(self):
|
||||
guide_path = Path(__file__).parent / "sdk" / "agent_generation_guide.md"
|
||||
content = guide_path.read_text(encoding="utf-8")
|
||||
clarify_section = content.split("Before or During Building")[1].split(
|
||||
"### Workflow"
|
||||
)[0]
|
||||
assert "ask_question" in clarify_section
|
||||
@@ -9,14 +9,11 @@ UTC). Fails open when Redis is unavailable to avoid blocking users.
|
||||
import asyncio
|
||||
import logging
|
||||
from datetime import UTC, datetime, timedelta
|
||||
from enum import Enum
|
||||
|
||||
from prisma.models import User as PrismaUser
|
||||
from pydantic import BaseModel, Field
|
||||
from redis.exceptions import RedisError
|
||||
|
||||
from backend.data.redis_client import get_redis_async
|
||||
from backend.util.cache import cached
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
@@ -24,40 +21,6 @@ logger = logging.getLogger(__name__)
|
||||
_USAGE_KEY_PREFIX = "copilot:usage"
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Subscription tier definitions
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class SubscriptionTier(str, Enum):
|
||||
"""Subscription tiers with increasing token allowances.
|
||||
|
||||
Mirrors the ``SubscriptionTier`` enum in ``schema.prisma``.
|
||||
Once ``prisma generate`` is run, this can be replaced with::
|
||||
|
||||
from prisma.enums import SubscriptionTier
|
||||
"""
|
||||
|
||||
FREE = "FREE"
|
||||
PRO = "PRO"
|
||||
BUSINESS = "BUSINESS"
|
||||
ENTERPRISE = "ENTERPRISE"
|
||||
|
||||
|
||||
# Multiplier applied to the base limits (from LD / config) for each tier.
|
||||
# Intentionally int (not float): keeps limits as whole token counts and avoids
|
||||
# floating-point rounding. If fractional multipliers are ever needed, change
|
||||
# the type and round the result in get_global_rate_limits().
|
||||
TIER_MULTIPLIERS: dict[SubscriptionTier, int] = {
|
||||
SubscriptionTier.FREE: 1,
|
||||
SubscriptionTier.PRO: 5,
|
||||
SubscriptionTier.BUSINESS: 20,
|
||||
SubscriptionTier.ENTERPRISE: 60,
|
||||
}
|
||||
|
||||
DEFAULT_TIER = SubscriptionTier.FREE
|
||||
|
||||
|
||||
class UsageWindow(BaseModel):
|
||||
"""Usage within a single time window."""
|
||||
|
||||
@@ -73,11 +36,6 @@ class CoPilotUsageStatus(BaseModel):
|
||||
|
||||
daily: UsageWindow
|
||||
weekly: UsageWindow
|
||||
tier: SubscriptionTier = DEFAULT_TIER
|
||||
reset_cost: int = Field(
|
||||
default=0,
|
||||
description="Credit cost (in cents) to reset the daily limit. 0 = feature disabled.",
|
||||
)
|
||||
|
||||
|
||||
class RateLimitExceeded(Exception):
|
||||
@@ -103,8 +61,6 @@ async def get_usage_status(
|
||||
user_id: str,
|
||||
daily_token_limit: int,
|
||||
weekly_token_limit: int,
|
||||
rate_limit_reset_cost: int = 0,
|
||||
tier: SubscriptionTier = DEFAULT_TIER,
|
||||
) -> CoPilotUsageStatus:
|
||||
"""Get current usage status for a user.
|
||||
|
||||
@@ -112,8 +68,6 @@ async def get_usage_status(
|
||||
user_id: The user's ID.
|
||||
daily_token_limit: Max tokens per day (0 = unlimited).
|
||||
weekly_token_limit: Max tokens per week (0 = unlimited).
|
||||
rate_limit_reset_cost: Credit cost (cents) to reset daily limit (0 = disabled).
|
||||
tier: The user's rate-limit tier (included in the response).
|
||||
|
||||
Returns:
|
||||
CoPilotUsageStatus with current usage and limits.
|
||||
@@ -143,8 +97,6 @@ async def get_usage_status(
|
||||
limit=weekly_token_limit,
|
||||
resets_at=_weekly_reset_time(now=now),
|
||||
),
|
||||
tier=tier,
|
||||
reset_cost=rate_limit_reset_cost,
|
||||
)
|
||||
|
||||
|
||||
@@ -189,111 +141,6 @@ async def check_rate_limit(
|
||||
raise RateLimitExceeded("weekly", _weekly_reset_time(now=now))
|
||||
|
||||
|
||||
async def reset_daily_usage(user_id: str, daily_token_limit: int = 0) -> bool:
|
||||
"""Reset a user's daily token usage counter in Redis.
|
||||
|
||||
Called after a user pays credits to extend their daily limit.
|
||||
Also reduces the weekly usage counter by ``daily_token_limit`` tokens
|
||||
(clamped to 0) so the user effectively gets one extra day's worth of
|
||||
weekly capacity.
|
||||
|
||||
Args:
|
||||
user_id: The user's ID.
|
||||
daily_token_limit: The configured daily token limit. When positive,
|
||||
the weekly counter is reduced by this amount.
|
||||
|
||||
Returns False if Redis is unavailable so the caller can handle
|
||||
compensation (fail-closed for billed operations, unlike the read-only
|
||||
rate-limit checks which fail-open).
|
||||
"""
|
||||
now = datetime.now(UTC)
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
|
||||
# Use a MULTI/EXEC transaction so that DELETE (daily) and DECRBY
|
||||
# (weekly) either both execute or neither does. This prevents the
|
||||
# scenario where the daily counter is cleared but the weekly
|
||||
# counter is not decremented — which would let the caller refund
|
||||
# credits even though the daily limit was already reset.
|
||||
d_key = _daily_key(user_id, now=now)
|
||||
w_key = _weekly_key(user_id, now=now) if daily_token_limit > 0 else None
|
||||
|
||||
pipe = redis.pipeline(transaction=True)
|
||||
pipe.delete(d_key)
|
||||
if w_key is not None:
|
||||
pipe.decrby(w_key, daily_token_limit)
|
||||
results = await pipe.execute()
|
||||
|
||||
# Clamp negative weekly counter to 0 (best-effort; not critical).
|
||||
if w_key is not None:
|
||||
new_val = results[1] # DECRBY result
|
||||
if new_val < 0:
|
||||
await redis.set(w_key, 0, keepttl=True)
|
||||
|
||||
logger.info("Reset daily usage for user %s", user_id[:8])
|
||||
return True
|
||||
except (RedisError, ConnectionError, OSError):
|
||||
logger.warning("Redis unavailable for resetting daily usage")
|
||||
return False
|
||||
|
||||
|
||||
_RESET_LOCK_PREFIX = "copilot:reset_lock"
|
||||
_RESET_COUNT_PREFIX = "copilot:reset_count"
|
||||
|
||||
|
||||
async def acquire_reset_lock(user_id: str, ttl_seconds: int = 10) -> bool:
|
||||
"""Acquire a short-lived lock to serialize rate limit resets per user."""
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
key = f"{_RESET_LOCK_PREFIX}:{user_id}"
|
||||
return bool(await redis.set(key, "1", nx=True, ex=ttl_seconds))
|
||||
except (RedisError, ConnectionError, OSError) as exc:
|
||||
logger.warning("Redis unavailable for reset lock, rejecting reset: %s", exc)
|
||||
return False
|
||||
|
||||
|
||||
async def release_reset_lock(user_id: str) -> None:
|
||||
"""Release the per-user reset lock."""
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
await redis.delete(f"{_RESET_LOCK_PREFIX}:{user_id}")
|
||||
except (RedisError, ConnectionError, OSError):
|
||||
pass # Lock will expire via TTL
|
||||
|
||||
|
||||
async def get_daily_reset_count(user_id: str) -> int | None:
|
||||
"""Get how many times the user has reset today.
|
||||
|
||||
Returns None when Redis is unavailable so callers can fail-closed
|
||||
for billed operations (as opposed to failing open for read-only
|
||||
rate-limit checks).
|
||||
"""
|
||||
now = datetime.now(UTC)
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
key = f"{_RESET_COUNT_PREFIX}:{user_id}:{now.strftime('%Y-%m-%d')}"
|
||||
val = await redis.get(key)
|
||||
return int(val or 0)
|
||||
except (RedisError, ConnectionError, OSError):
|
||||
logger.warning("Redis unavailable for reading daily reset count")
|
||||
return None
|
||||
|
||||
|
||||
async def increment_daily_reset_count(user_id: str) -> None:
|
||||
"""Increment and track how many resets this user has done today."""
|
||||
now = datetime.now(UTC)
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
key = f"{_RESET_COUNT_PREFIX}:{user_id}:{now.strftime('%Y-%m-%d')}"
|
||||
pipe = redis.pipeline(transaction=True)
|
||||
pipe.incr(key)
|
||||
seconds_until_reset = int((_daily_reset_time(now=now) - now).total_seconds())
|
||||
pipe.expire(key, max(seconds_until_reset, 1))
|
||||
await pipe.execute()
|
||||
except (RedisError, ConnectionError, OSError):
|
||||
logger.warning("Redis unavailable for tracking reset count")
|
||||
|
||||
|
||||
async def record_token_usage(
|
||||
user_id: str,
|
||||
prompt_tokens: int,
|
||||
@@ -384,155 +231,6 @@ async def record_token_usage(
|
||||
)
|
||||
|
||||
|
||||
class _UserNotFoundError(Exception):
|
||||
"""Raised when a user record is missing or has no subscription tier.
|
||||
|
||||
Used internally by ``_fetch_user_tier`` to signal a cache-miss condition:
|
||||
by raising instead of returning ``DEFAULT_TIER``, we prevent the ``@cached``
|
||||
decorator from storing the fallback value. This avoids a race condition
|
||||
where a non-existent user's DEFAULT_TIER is cached, then the user is
|
||||
created with a higher tier but receives the stale cached FREE tier for
|
||||
up to 5 minutes.
|
||||
"""
|
||||
|
||||
|
||||
@cached(maxsize=1000, ttl_seconds=300, shared_cache=True)
|
||||
async def _fetch_user_tier(user_id: str) -> SubscriptionTier:
|
||||
"""Fetch the user's rate-limit tier from the database (cached via Redis).
|
||||
|
||||
Uses ``shared_cache=True`` so that tier changes propagate across all pods
|
||||
immediately when the cache entry is invalidated (via ``cache_delete``).
|
||||
|
||||
Only successful DB lookups of existing users with a valid tier are cached.
|
||||
Raises ``_UserNotFoundError`` when the user is missing or has no tier, so
|
||||
the ``@cached`` decorator does **not** store a fallback value. This
|
||||
prevents a race condition where a non-existent user's ``DEFAULT_TIER`` is
|
||||
cached and then persists after the user is created with a higher tier.
|
||||
"""
|
||||
user = await PrismaUser.prisma().find_unique(where={"id": user_id})
|
||||
if user and user.subscriptionTier: # type: ignore[reportAttributeAccessIssue]
|
||||
return SubscriptionTier(user.subscriptionTier) # type: ignore[reportAttributeAccessIssue]
|
||||
raise _UserNotFoundError(user_id)
|
||||
|
||||
|
||||
async def get_user_tier(user_id: str) -> SubscriptionTier:
|
||||
"""Look up the user's rate-limit tier from the database.
|
||||
|
||||
Successful results are cached for 5 minutes (via ``_fetch_user_tier``)
|
||||
to avoid a DB round-trip on every rate-limit check.
|
||||
|
||||
Falls back to ``DEFAULT_TIER`` **without caching** when the DB is
|
||||
unreachable or returns an unrecognised value, so the next call retries
|
||||
the query instead of serving a stale fallback for up to 5 minutes.
|
||||
"""
|
||||
try:
|
||||
return await _fetch_user_tier(user_id)
|
||||
except Exception as exc:
|
||||
logger.warning(
|
||||
"Failed to resolve rate-limit tier for user %s, defaulting to %s: %s",
|
||||
user_id[:8],
|
||||
DEFAULT_TIER.value,
|
||||
exc,
|
||||
)
|
||||
return DEFAULT_TIER
|
||||
|
||||
|
||||
# Expose cache management on the public function so callers (including tests)
|
||||
# never need to reach into the private ``_fetch_user_tier``.
|
||||
get_user_tier.cache_clear = _fetch_user_tier.cache_clear # type: ignore[attr-defined]
|
||||
get_user_tier.cache_delete = _fetch_user_tier.cache_delete # type: ignore[attr-defined]
|
||||
|
||||
|
||||
async def set_user_tier(user_id: str, tier: SubscriptionTier) -> None:
|
||||
"""Persist the user's rate-limit tier to the database.
|
||||
|
||||
Also invalidates the ``get_user_tier`` cache for this user so that
|
||||
subsequent rate-limit checks immediately see the new tier.
|
||||
|
||||
Raises:
|
||||
prisma.errors.RecordNotFoundError: If the user does not exist.
|
||||
"""
|
||||
await PrismaUser.prisma().update(
|
||||
where={"id": user_id},
|
||||
data={"subscriptionTier": tier.value},
|
||||
)
|
||||
# Invalidate cached tier so rate-limit checks pick up the change immediately.
|
||||
get_user_tier.cache_delete(user_id) # type: ignore[attr-defined]
|
||||
|
||||
|
||||
async def get_global_rate_limits(
|
||||
user_id: str,
|
||||
config_daily: int,
|
||||
config_weekly: int,
|
||||
) -> tuple[int, int, SubscriptionTier]:
|
||||
"""Resolve global rate limits from LaunchDarkly, falling back to config.
|
||||
|
||||
The base limits (from LD or config) are multiplied by the user's
|
||||
tier multiplier so that higher tiers receive proportionally larger
|
||||
allowances.
|
||||
|
||||
Args:
|
||||
user_id: User ID for LD flag evaluation context.
|
||||
config_daily: Fallback daily limit from ChatConfig.
|
||||
config_weekly: Fallback weekly limit from ChatConfig.
|
||||
|
||||
Returns:
|
||||
(daily_token_limit, weekly_token_limit, tier) 3-tuple.
|
||||
"""
|
||||
# Lazy import to avoid circular dependency:
|
||||
# rate_limit -> feature_flag -> settings -> ... -> rate_limit
|
||||
from backend.util.feature_flag import Flag, get_feature_flag_value
|
||||
|
||||
daily_raw = await get_feature_flag_value(
|
||||
Flag.COPILOT_DAILY_TOKEN_LIMIT.value, user_id, config_daily
|
||||
)
|
||||
weekly_raw = await get_feature_flag_value(
|
||||
Flag.COPILOT_WEEKLY_TOKEN_LIMIT.value, user_id, config_weekly
|
||||
)
|
||||
try:
|
||||
daily = max(0, int(daily_raw))
|
||||
except (TypeError, ValueError):
|
||||
logger.warning("Invalid LD value for daily token limit: %r", daily_raw)
|
||||
daily = config_daily
|
||||
try:
|
||||
weekly = max(0, int(weekly_raw))
|
||||
except (TypeError, ValueError):
|
||||
logger.warning("Invalid LD value for weekly token limit: %r", weekly_raw)
|
||||
weekly = config_weekly
|
||||
|
||||
# Apply tier multiplier
|
||||
tier = await get_user_tier(user_id)
|
||||
multiplier = TIER_MULTIPLIERS.get(tier, 1)
|
||||
if multiplier != 1:
|
||||
daily = daily * multiplier
|
||||
weekly = weekly * multiplier
|
||||
|
||||
return daily, weekly, tier
|
||||
|
||||
|
||||
async def reset_user_usage(user_id: str, *, reset_weekly: bool = False) -> None:
|
||||
"""Reset a user's usage counters.
|
||||
|
||||
Always deletes the daily Redis key. When *reset_weekly* is ``True``,
|
||||
the weekly key is deleted as well.
|
||||
|
||||
Unlike read paths (``get_usage_status``, ``check_rate_limit``) which
|
||||
fail-open on Redis errors, resets intentionally re-raise so the caller
|
||||
knows the operation did not succeed. A silent failure here would leave
|
||||
the admin believing the counters were zeroed when they were not.
|
||||
"""
|
||||
now = datetime.now(UTC)
|
||||
keys_to_delete = [_daily_key(user_id, now=now)]
|
||||
if reset_weekly:
|
||||
keys_to_delete.append(_weekly_key(user_id, now=now))
|
||||
try:
|
||||
redis = await get_redis_async()
|
||||
await redis.delete(*keys_to_delete)
|
||||
except (RedisError, ConnectionError, OSError):
|
||||
logger.warning("Redis unavailable for resetting user usage")
|
||||
raise
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Private helpers
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
@@ -7,19 +7,11 @@ import pytest
|
||||
from redis.exceptions import RedisError
|
||||
|
||||
from .rate_limit import (
|
||||
DEFAULT_TIER,
|
||||
TIER_MULTIPLIERS,
|
||||
CoPilotUsageStatus,
|
||||
RateLimitExceeded,
|
||||
SubscriptionTier,
|
||||
UsageWindow,
|
||||
check_rate_limit,
|
||||
get_global_rate_limits,
|
||||
get_usage_status,
|
||||
get_user_tier,
|
||||
record_token_usage,
|
||||
reset_daily_usage,
|
||||
set_user_tier,
|
||||
)
|
||||
|
||||
_USER = "test-user-rl"
|
||||
@@ -340,873 +332,3 @@ class TestRecordTokenUsage:
|
||||
):
|
||||
# Should not raise — fail-open
|
||||
await record_token_usage(_USER, prompt_tokens=100, completion_tokens=50)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# SubscriptionTier and tier multipliers
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestSubscriptionTier:
|
||||
def test_tier_values(self):
|
||||
assert SubscriptionTier.FREE.value == "FREE"
|
||||
assert SubscriptionTier.PRO.value == "PRO"
|
||||
assert SubscriptionTier.BUSINESS.value == "BUSINESS"
|
||||
assert SubscriptionTier.ENTERPRISE.value == "ENTERPRISE"
|
||||
|
||||
def test_tier_multipliers(self):
|
||||
assert TIER_MULTIPLIERS[SubscriptionTier.FREE] == 1
|
||||
assert TIER_MULTIPLIERS[SubscriptionTier.PRO] == 5
|
||||
assert TIER_MULTIPLIERS[SubscriptionTier.BUSINESS] == 20
|
||||
assert TIER_MULTIPLIERS[SubscriptionTier.ENTERPRISE] == 60
|
||||
|
||||
def test_default_tier_is_free(self):
|
||||
assert DEFAULT_TIER == SubscriptionTier.FREE
|
||||
|
||||
def test_usage_status_includes_tier(self):
|
||||
now = datetime.now(UTC)
|
||||
status = CoPilotUsageStatus(
|
||||
daily=UsageWindow(used=0, limit=100, resets_at=now + timedelta(hours=1)),
|
||||
weekly=UsageWindow(used=0, limit=500, resets_at=now + timedelta(days=1)),
|
||||
)
|
||||
assert status.tier == SubscriptionTier.FREE
|
||||
|
||||
def test_usage_status_with_custom_tier(self):
|
||||
now = datetime.now(UTC)
|
||||
status = CoPilotUsageStatus(
|
||||
daily=UsageWindow(used=0, limit=100, resets_at=now + timedelta(hours=1)),
|
||||
weekly=UsageWindow(used=0, limit=500, resets_at=now + timedelta(days=1)),
|
||||
tier=SubscriptionTier.PRO,
|
||||
)
|
||||
assert status.tier == SubscriptionTier.PRO
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# get_user_tier
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestGetUserTier:
|
||||
@pytest.fixture(autouse=True)
|
||||
def _clear_tier_cache(self):
|
||||
"""Clear the get_user_tier cache before each test."""
|
||||
get_user_tier.cache_clear() # type: ignore[attr-defined]
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_tier_from_db(self):
|
||||
"""Should return the tier stored in the user record."""
|
||||
mock_user = MagicMock()
|
||||
mock_user.subscriptionTier = "PRO"
|
||||
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.find_unique = AsyncMock(return_value=mock_user)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
tier = await get_user_tier(_USER)
|
||||
|
||||
assert tier == SubscriptionTier.PRO
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_default_when_user_not_found(self):
|
||||
"""Should return DEFAULT_TIER when user is not in the DB."""
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.find_unique = AsyncMock(return_value=None)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
tier = await get_user_tier(_USER)
|
||||
|
||||
assert tier == DEFAULT_TIER
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_default_when_tier_is_none(self):
|
||||
"""Should return DEFAULT_TIER when subscriptionTier is None."""
|
||||
mock_user = MagicMock()
|
||||
mock_user.subscriptionTier = None
|
||||
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.find_unique = AsyncMock(return_value=mock_user)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
tier = await get_user_tier(_USER)
|
||||
|
||||
assert tier == DEFAULT_TIER
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_default_on_db_error(self):
|
||||
"""Should fall back to DEFAULT_TIER when DB raises."""
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.find_unique = AsyncMock(side_effect=Exception("DB down"))
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
tier = await get_user_tier(_USER)
|
||||
|
||||
assert tier == DEFAULT_TIER
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_db_error_is_not_cached(self):
|
||||
"""Transient DB errors should NOT cache the default tier.
|
||||
|
||||
Regression test: a transient DB failure previously cached DEFAULT_TIER
|
||||
for 5 minutes, incorrectly downgrading higher-tier users until expiry.
|
||||
"""
|
||||
failing_prisma = AsyncMock()
|
||||
failing_prisma.find_unique = AsyncMock(side_effect=Exception("DB down"))
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=failing_prisma,
|
||||
):
|
||||
tier1 = await get_user_tier(_USER)
|
||||
assert tier1 == DEFAULT_TIER
|
||||
|
||||
# Now DB recovers and returns PRO
|
||||
mock_user = MagicMock()
|
||||
mock_user.subscriptionTier = "PRO"
|
||||
ok_prisma = AsyncMock()
|
||||
ok_prisma.find_unique = AsyncMock(return_value=mock_user)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=ok_prisma,
|
||||
):
|
||||
tier2 = await get_user_tier(_USER)
|
||||
|
||||
# Should get PRO now — the error result was not cached
|
||||
assert tier2 == SubscriptionTier.PRO
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_default_on_invalid_tier_value(self):
|
||||
"""Should fall back to DEFAULT_TIER when stored value is invalid."""
|
||||
mock_user = MagicMock()
|
||||
mock_user.subscriptionTier = "invalid-tier"
|
||||
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.find_unique = AsyncMock(return_value=mock_user)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
tier = await get_user_tier(_USER)
|
||||
|
||||
assert tier == DEFAULT_TIER
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_user_not_found_is_not_cached(self):
|
||||
"""Non-existent user should NOT cache DEFAULT_TIER.
|
||||
|
||||
Regression test: when ``get_user_tier`` is called before a user record
|
||||
exists, the DEFAULT_TIER fallback must not be cached. Otherwise, a
|
||||
newly created user with a higher tier (e.g. PRO) would receive the
|
||||
stale cached FREE tier for up to 5 minutes.
|
||||
"""
|
||||
# First call: user does not exist yet
|
||||
missing_prisma = AsyncMock()
|
||||
missing_prisma.find_unique = AsyncMock(return_value=None)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=missing_prisma,
|
||||
):
|
||||
tier1 = await get_user_tier(_USER)
|
||||
assert tier1 == DEFAULT_TIER
|
||||
|
||||
# Second call: user now exists with PRO tier
|
||||
mock_user = MagicMock()
|
||||
mock_user.subscriptionTier = "PRO"
|
||||
ok_prisma = AsyncMock()
|
||||
ok_prisma.find_unique = AsyncMock(return_value=mock_user)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=ok_prisma,
|
||||
):
|
||||
tier2 = await get_user_tier(_USER)
|
||||
|
||||
# Should get PRO — the not-found result was not cached
|
||||
assert tier2 == SubscriptionTier.PRO
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# set_user_tier
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestSetUserTier:
|
||||
@pytest.fixture(autouse=True)
|
||||
def _clear_tier_cache(self):
|
||||
"""Clear the get_user_tier cache before each test."""
|
||||
get_user_tier.cache_clear() # type: ignore[attr-defined]
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_updates_db_and_invalidates_cache(self):
|
||||
"""set_user_tier should persist to DB and invalidate the tier cache."""
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.update = AsyncMock(return_value=None)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
await set_user_tier(_USER, SubscriptionTier.PRO)
|
||||
|
||||
mock_prisma.update.assert_awaited_once_with(
|
||||
where={"id": _USER},
|
||||
data={"subscriptionTier": "PRO"},
|
||||
)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_record_not_found_propagates(self):
|
||||
"""RecordNotFoundError from Prisma should propagate to callers."""
|
||||
import prisma.errors
|
||||
|
||||
mock_prisma = AsyncMock()
|
||||
mock_prisma.update = AsyncMock(
|
||||
side_effect=prisma.errors.RecordNotFoundError(
|
||||
{"error": "Record not found"}
|
||||
),
|
||||
)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma,
|
||||
):
|
||||
with pytest.raises(prisma.errors.RecordNotFoundError):
|
||||
await set_user_tier(_USER, SubscriptionTier.ENTERPRISE)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_cache_invalidated_after_set(self):
|
||||
"""After set_user_tier, get_user_tier should query DB again (not cache)."""
|
||||
# First, populate the cache with BUSINESS
|
||||
mock_user_biz = MagicMock()
|
||||
mock_user_biz.subscriptionTier = "BUSINESS"
|
||||
mock_prisma_get = AsyncMock()
|
||||
mock_prisma_get.find_unique = AsyncMock(return_value=mock_user_biz)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma_get,
|
||||
):
|
||||
tier_before = await get_user_tier(_USER)
|
||||
assert tier_before == SubscriptionTier.BUSINESS
|
||||
|
||||
# Now set tier to ENTERPRISE (this should invalidate the cache)
|
||||
mock_prisma_set = AsyncMock()
|
||||
mock_prisma_set.update = AsyncMock(return_value=None)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma_set,
|
||||
):
|
||||
await set_user_tier(_USER, SubscriptionTier.ENTERPRISE)
|
||||
|
||||
# Now get_user_tier should hit DB again (cache was invalidated)
|
||||
mock_user_ent = MagicMock()
|
||||
mock_user_ent.subscriptionTier = "ENTERPRISE"
|
||||
mock_prisma_get2 = AsyncMock()
|
||||
mock_prisma_get2.find_unique = AsyncMock(return_value=mock_user_ent)
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.PrismaUser.prisma",
|
||||
return_value=mock_prisma_get2,
|
||||
):
|
||||
tier_after = await get_user_tier(_USER)
|
||||
|
||||
assert tier_after == SubscriptionTier.ENTERPRISE
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# get_global_rate_limits with tiers
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestGetGlobalRateLimitsWithTiers:
|
||||
@staticmethod
|
||||
def _ld_side_effect(daily: int, weekly: int):
|
||||
"""Return an async side_effect that dispatches by flag_key."""
|
||||
|
||||
async def _side_effect(flag_key: str, _uid: str, default: int) -> int:
|
||||
if "daily" in flag_key.lower():
|
||||
return daily
|
||||
if "weekly" in flag_key.lower():
|
||||
return weekly
|
||||
return default
|
||||
|
||||
return _side_effect
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_free_tier_no_multiplier(self):
|
||||
"""Free tier should not change limits."""
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.FREE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(2_500_000, 12_500_000),
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, 2_500_000, 12_500_000
|
||||
)
|
||||
|
||||
assert daily == 2_500_000
|
||||
assert weekly == 12_500_000
|
||||
assert tier == SubscriptionTier.FREE
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_pro_tier_5x_multiplier(self):
|
||||
"""Pro tier should multiply limits by 5."""
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(2_500_000, 12_500_000),
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, 2_500_000, 12_500_000
|
||||
)
|
||||
|
||||
assert daily == 12_500_000
|
||||
assert weekly == 62_500_000
|
||||
assert tier == SubscriptionTier.PRO
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_business_tier_20x_multiplier(self):
|
||||
"""Business tier should multiply limits by 20."""
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.BUSINESS,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(2_500_000, 12_500_000),
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, 2_500_000, 12_500_000
|
||||
)
|
||||
|
||||
assert daily == 50_000_000
|
||||
assert weekly == 250_000_000
|
||||
assert tier == SubscriptionTier.BUSINESS
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_enterprise_tier_60x_multiplier(self):
|
||||
"""Enterprise tier should multiply limits by 60."""
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.ENTERPRISE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(2_500_000, 12_500_000),
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, 2_500_000, 12_500_000
|
||||
)
|
||||
|
||||
assert daily == 150_000_000
|
||||
assert weekly == 750_000_000
|
||||
assert tier == SubscriptionTier.ENTERPRISE
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# End-to-end: tier limits are respected by check_rate_limit
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestTierLimitsRespected:
|
||||
"""Verify that tier-adjusted limits from get_global_rate_limits flow
|
||||
correctly into check_rate_limit, so higher tiers allow more usage and
|
||||
lower tiers are blocked when they would exceed their allocation."""
|
||||
|
||||
_BASE_DAILY = 2_500_000
|
||||
_BASE_WEEKLY = 12_500_000
|
||||
|
||||
@staticmethod
|
||||
def _ld_side_effect(daily: int, weekly: int):
|
||||
|
||||
async def _side_effect(flag_key: str, _uid: str, default: int) -> int:
|
||||
if "daily" in flag_key.lower():
|
||||
return daily
|
||||
if "weekly" in flag_key.lower():
|
||||
return weekly
|
||||
return default
|
||||
|
||||
return _side_effect
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_pro_user_allowed_above_free_limit(self):
|
||||
"""A PRO user with usage above the FREE limit should be allowed."""
|
||||
# Usage: 3M tokens (above FREE limit of 2.5M, below PRO limit of 12.5M)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=["3000000", "3000000"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
# PRO: 5x multiplier
|
||||
assert daily == 12_500_000
|
||||
assert tier == SubscriptionTier.PRO
|
||||
# Should NOT raise — 3M < 12.5M
|
||||
await check_rate_limit(
|
||||
_USER, daily_token_limit=daily, weekly_token_limit=weekly
|
||||
)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_free_user_blocked_at_free_limit(self):
|
||||
"""A FREE user at or above the base limit should be blocked."""
|
||||
# Usage: 2.5M tokens (at FREE limit of 2.5M)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=["2500000", "2500000"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.FREE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
# FREE: 1x multiplier
|
||||
assert daily == 2_500_000
|
||||
assert tier == SubscriptionTier.FREE
|
||||
# Should raise — 2.5M >= 2.5M
|
||||
with pytest.raises(RateLimitExceeded):
|
||||
await check_rate_limit(
|
||||
_USER, daily_token_limit=daily, weekly_token_limit=weekly
|
||||
)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_enterprise_user_has_highest_headroom(self):
|
||||
"""An ENTERPRISE user should have 60x the base limit."""
|
||||
# Usage: 100M tokens (huge, but below ENTERPRISE daily of 150M)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=["100000000", "100000000"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.ENTERPRISE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert daily == 150_000_000
|
||||
assert tier == SubscriptionTier.ENTERPRISE
|
||||
# Should NOT raise — 100M < 150M
|
||||
await check_rate_limit(
|
||||
_USER, daily_token_limit=daily, weekly_token_limit=weekly
|
||||
)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# reset_daily_usage
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestResetDailyUsage:
|
||||
@staticmethod
|
||||
def _make_pipeline_mock(decrby_result: int = 0) -> MagicMock:
|
||||
"""Create a pipeline mock that returns [delete_result, decrby_result]."""
|
||||
pipe = MagicMock()
|
||||
pipe.execute = AsyncMock(return_value=[1, decrby_result])
|
||||
return pipe
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_deletes_daily_key(self):
|
||||
mock_pipe = self._make_pipeline_mock(decrby_result=0)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.pipeline = lambda **_kw: mock_pipe
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
result = await reset_daily_usage(_USER, daily_token_limit=10000)
|
||||
|
||||
assert result is True
|
||||
mock_pipe.delete.assert_called_once()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_reduces_weekly_usage_via_decrby(self):
|
||||
"""Weekly counter should be reduced via DECRBY in the pipeline."""
|
||||
mock_pipe = self._make_pipeline_mock(decrby_result=35000)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.pipeline = lambda **_kw: mock_pipe
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
await reset_daily_usage(_USER, daily_token_limit=10000)
|
||||
|
||||
mock_pipe.decrby.assert_called_once()
|
||||
mock_redis.set.assert_not_called() # 35000 > 0, no clamp needed
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_clamps_negative_weekly_to_zero(self):
|
||||
"""If DECRBY goes negative, SET to 0 (outside the pipeline)."""
|
||||
mock_pipe = self._make_pipeline_mock(decrby_result=-5000)
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.pipeline = lambda **_kw: mock_pipe
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
await reset_daily_usage(_USER, daily_token_limit=10000)
|
||||
|
||||
mock_pipe.decrby.assert_called_once()
|
||||
mock_redis.set.assert_called_once()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_no_weekly_reduction_when_daily_limit_zero(self):
|
||||
"""When daily_token_limit is 0, weekly counter should not be touched."""
|
||||
mock_pipe = self._make_pipeline_mock()
|
||||
mock_pipe.execute = AsyncMock(return_value=[1]) # only delete result
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.pipeline = lambda **_kw: mock_pipe
|
||||
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
):
|
||||
await reset_daily_usage(_USER, daily_token_limit=0)
|
||||
|
||||
mock_pipe.delete.assert_called_once()
|
||||
mock_pipe.decrby.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_false_when_redis_unavailable(self):
|
||||
with patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
side_effect=ConnectionError("Redis down"),
|
||||
):
|
||||
result = await reset_daily_usage(_USER, daily_token_limit=10000)
|
||||
|
||||
assert result is False
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Tier-limit enforcement (integration-style)
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestTierLimitsEnforced:
|
||||
"""Verify that tier-multiplied limits are actually respected by
|
||||
``check_rate_limit`` — i.e. that usage within the tier allowance passes
|
||||
and usage at/above the tier allowance is rejected."""
|
||||
|
||||
_BASE_DAILY = 1_000_000
|
||||
_BASE_WEEKLY = 5_000_000
|
||||
|
||||
@staticmethod
|
||||
def _ld_side_effect(daily: int, weekly: int):
|
||||
"""Mock LD flag lookup returning the given raw limits."""
|
||||
|
||||
async def _side_effect(flag_key: str, _uid: str, default: int) -> int:
|
||||
if "daily" in flag_key.lower():
|
||||
return daily
|
||||
if "weekly" in flag_key.lower():
|
||||
return weekly
|
||||
return default
|
||||
|
||||
return _side_effect
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_pro_within_limit_allowed(self):
|
||||
"""Usage under PRO daily limit should not raise."""
|
||||
pro_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.PRO]
|
||||
mock_redis = AsyncMock()
|
||||
# Simulate usage just under the PRO daily limit
|
||||
mock_redis.get = AsyncMock(side_effect=[str(pro_daily - 1), "0"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert tier == SubscriptionTier.PRO
|
||||
assert daily == pro_daily
|
||||
# Should not raise — usage is under the limit
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_pro_at_limit_rejected(self):
|
||||
"""Usage at exactly the PRO daily limit should raise."""
|
||||
pro_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.PRO]
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=[str(pro_daily), "0"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
with pytest.raises(RateLimitExceeded) as exc_info:
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
assert exc_info.value.window == "daily"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_business_higher_limit_allows_pro_overflow(self):
|
||||
"""Usage exceeding PRO but under BUSINESS should pass for BUSINESS."""
|
||||
pro_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.PRO]
|
||||
biz_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.BUSINESS]
|
||||
# Usage between PRO and BUSINESS limits
|
||||
usage = pro_daily + 1_000_000
|
||||
assert usage < biz_daily, "test sanity: usage must be under BUSINESS limit"
|
||||
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=[str(usage), "0"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.BUSINESS,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert tier == SubscriptionTier.BUSINESS
|
||||
assert daily == biz_daily
|
||||
# Should not raise — BUSINESS tier can handle this
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_weekly_limit_enforced_for_tier(self):
|
||||
"""Weekly limit should also be tier-multiplied and enforced."""
|
||||
pro_weekly = self._BASE_WEEKLY * TIER_MULTIPLIERS[SubscriptionTier.PRO]
|
||||
mock_redis = AsyncMock()
|
||||
# Daily usage fine, weekly at limit
|
||||
mock_redis.get = AsyncMock(side_effect=["0", str(pro_weekly)])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.PRO,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
with pytest.raises(RateLimitExceeded) as exc_info:
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
assert exc_info.value.window == "weekly"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_free_tier_base_limit_enforced(self):
|
||||
"""Free tier (1x multiplier) should enforce the base limit exactly."""
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=[str(self._BASE_DAILY), "0"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.FREE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert daily == self._BASE_DAILY # 1x multiplier
|
||||
with pytest.raises(RateLimitExceeded):
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_free_tier_cannot_bypass_pro_limit(self):
|
||||
"""A FREE-tier user whose usage is within PRO limits but over FREE
|
||||
limits must still be rejected.
|
||||
|
||||
Negative test: ensures the tier multiplier is applied *before* the
|
||||
rate-limit check, so a lower-tier user cannot 'bypass' limits that
|
||||
would be acceptable for a higher tier.
|
||||
"""
|
||||
free_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.FREE]
|
||||
pro_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.PRO]
|
||||
# Usage above FREE limit but below PRO limit
|
||||
usage = free_daily + 500_000
|
||||
assert usage < pro_daily, "test sanity: usage must be under PRO limit"
|
||||
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=[str(usage), "0"])
|
||||
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.FREE,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert tier == SubscriptionTier.FREE
|
||||
assert daily == free_daily # 1x, not 5x
|
||||
with pytest.raises(RateLimitExceeded) as exc_info:
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
assert exc_info.value.window == "daily"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_tier_change_updates_effective_limits(self):
|
||||
"""After upgrading from FREE to BUSINESS, the effective limits must
|
||||
increase accordingly.
|
||||
|
||||
Verifies that the tier multiplier is correctly applied after a tier
|
||||
change, and that usage that was over the FREE limit is within the new
|
||||
BUSINESS limit.
|
||||
"""
|
||||
free_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.FREE]
|
||||
biz_daily = self._BASE_DAILY * TIER_MULTIPLIERS[SubscriptionTier.BUSINESS]
|
||||
# Usage above FREE limit but below BUSINESS limit
|
||||
usage = free_daily + 500_000
|
||||
assert usage < biz_daily, "test sanity: usage must be under BUSINESS limit"
|
||||
|
||||
mock_redis = AsyncMock()
|
||||
mock_redis.get = AsyncMock(side_effect=[str(usage), "0"])
|
||||
|
||||
# Simulate the user having been upgraded to BUSINESS
|
||||
with (
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_user_tier",
|
||||
new_callable=AsyncMock,
|
||||
return_value=SubscriptionTier.BUSINESS,
|
||||
),
|
||||
patch(
|
||||
"backend.util.feature_flag.get_feature_flag_value",
|
||||
side_effect=self._ld_side_effect(self._BASE_DAILY, self._BASE_WEEKLY),
|
||||
),
|
||||
patch(
|
||||
"backend.copilot.rate_limit.get_redis_async",
|
||||
return_value=mock_redis,
|
||||
),
|
||||
):
|
||||
daily, weekly, tier = await get_global_rate_limits(
|
||||
_USER, self._BASE_DAILY, self._BASE_WEEKLY
|
||||
)
|
||||
assert tier == SubscriptionTier.BUSINESS
|
||||
assert daily == biz_daily # 20x
|
||||
# Should NOT raise — usage is within the BUSINESS tier allowance
|
||||
await check_rate_limit(_USER, daily, weekly)
|
||||
|
||||
@@ -1,315 +0,0 @@
|
||||
"""Unit tests for the POST /usage/reset endpoint."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from datetime import UTC, datetime, timedelta
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
|
||||
import pytest
|
||||
from fastapi import HTTPException
|
||||
|
||||
from backend.api.features.chat.routes import reset_copilot_usage
|
||||
from backend.copilot.rate_limit import CoPilotUsageStatus, SubscriptionTier, UsageWindow
|
||||
from backend.util.exceptions import InsufficientBalanceError
|
||||
|
||||
|
||||
# Minimal config mock matching ChatConfig fields used by the endpoint.
|
||||
def _make_config(
|
||||
rate_limit_reset_cost: int = 500,
|
||||
daily_token_limit: int = 2_500_000,
|
||||
weekly_token_limit: int = 12_500_000,
|
||||
max_daily_resets: int = 5,
|
||||
):
|
||||
cfg = MagicMock()
|
||||
cfg.rate_limit_reset_cost = rate_limit_reset_cost
|
||||
cfg.daily_token_limit = daily_token_limit
|
||||
cfg.weekly_token_limit = weekly_token_limit
|
||||
cfg.max_daily_resets = max_daily_resets
|
||||
return cfg
|
||||
|
||||
|
||||
def _usage(daily_used: int = 3_000_000, daily_limit: int = 2_500_000):
|
||||
return CoPilotUsageStatus(
|
||||
daily=UsageWindow(
|
||||
used=daily_used,
|
||||
limit=daily_limit,
|
||||
resets_at=datetime.now(UTC) + timedelta(hours=6),
|
||||
),
|
||||
weekly=UsageWindow(
|
||||
used=5_000_000,
|
||||
limit=12_500_000,
|
||||
resets_at=datetime.now(UTC) + timedelta(days=3),
|
||||
),
|
||||
)
|
||||
|
||||
|
||||
_MODULE = "backend.api.features.chat.routes"
|
||||
|
||||
|
||||
def _mock_settings(enable_credit: bool = True):
|
||||
"""Return a mock Settings object with the given enable_credit flag."""
|
||||
mock = MagicMock()
|
||||
mock.config.enable_credit = enable_credit
|
||||
return mock
|
||||
|
||||
|
||||
def _mock_rate_limits(
|
||||
daily: int = 2_500_000,
|
||||
weekly: int = 12_500_000,
|
||||
tier: SubscriptionTier = SubscriptionTier.PRO,
|
||||
):
|
||||
"""Mock get_global_rate_limits to return fixed limits (no tier multiplier)."""
|
||||
return patch(
|
||||
f"{_MODULE}.get_global_rate_limits",
|
||||
AsyncMock(return_value=(daily, weekly, tier)),
|
||||
)
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
class TestResetCopilotUsage:
|
||||
async def test_feature_disabled_returns_400(self):
|
||||
"""When rate_limit_reset_cost=0, endpoint returns 400."""
|
||||
|
||||
with patch(f"{_MODULE}.config", _make_config(rate_limit_reset_cost=0)):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 400
|
||||
assert "not available" in exc_info.value.detail
|
||||
|
||||
async def test_no_daily_limit_returns_400(self):
|
||||
"""When daily_token_limit=0 (unlimited), endpoint returns 400."""
|
||||
|
||||
with (
|
||||
patch(f"{_MODULE}.config", _make_config(daily_token_limit=0)),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(daily=0),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 400
|
||||
assert "nothing to reset" in exc_info.value.detail.lower()
|
||||
|
||||
async def test_not_at_limit_returns_400(self):
|
||||
"""When user hasn't hit their daily limit, returns 400."""
|
||||
|
||||
cfg = _make_config()
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()) as mock_release,
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(return_value=_usage(daily_used=1_000_000)),
|
||||
),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 400
|
||||
assert "not reached" in exc_info.value.detail
|
||||
mock_release.assert_awaited_once()
|
||||
|
||||
async def test_insufficient_credits_returns_402(self):
|
||||
"""When user doesn't have enough credits, returns 402."""
|
||||
|
||||
mock_credit_model = AsyncMock()
|
||||
mock_credit_model.spend_credits.side_effect = InsufficientBalanceError(
|
||||
message="Insufficient balance",
|
||||
user_id="user-1",
|
||||
balance=50,
|
||||
amount=200,
|
||||
)
|
||||
|
||||
cfg = _make_config()
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()) as mock_release,
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(return_value=_usage()),
|
||||
),
|
||||
patch(
|
||||
f"{_MODULE}.get_user_credit_model",
|
||||
AsyncMock(return_value=mock_credit_model),
|
||||
),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 402
|
||||
mock_release.assert_awaited_once()
|
||||
|
||||
async def test_happy_path(self):
|
||||
"""Successful reset: charges credits, resets usage, returns response."""
|
||||
|
||||
mock_credit_model = AsyncMock()
|
||||
mock_credit_model.spend_credits.return_value = 1500 # remaining balance
|
||||
|
||||
cfg = _make_config()
|
||||
updated_usage = _usage(daily_used=0)
|
||||
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()),
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(side_effect=[_usage(), updated_usage]),
|
||||
),
|
||||
patch(
|
||||
f"{_MODULE}.get_user_credit_model",
|
||||
AsyncMock(return_value=mock_credit_model),
|
||||
),
|
||||
patch(
|
||||
f"{_MODULE}.reset_daily_usage", AsyncMock(return_value=True)
|
||||
) as mock_reset,
|
||||
patch(f"{_MODULE}.increment_daily_reset_count", AsyncMock()) as mock_incr,
|
||||
):
|
||||
result = await reset_copilot_usage(user_id="user-1")
|
||||
assert result.success is True
|
||||
assert result.credits_charged == 500
|
||||
assert result.remaining_balance == 1500
|
||||
mock_reset.assert_awaited_once()
|
||||
mock_incr.assert_awaited_once()
|
||||
|
||||
async def test_max_daily_resets_exceeded(self):
|
||||
"""When user has exhausted daily resets, returns 429."""
|
||||
|
||||
cfg = _make_config(max_daily_resets=3)
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=3)),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 429
|
||||
|
||||
async def test_credit_system_disabled_returns_400(self):
|
||||
"""When enable_credit=False, endpoint returns 400."""
|
||||
|
||||
with (
|
||||
patch(f"{_MODULE}.config", _make_config()),
|
||||
patch(f"{_MODULE}.settings", _mock_settings(enable_credit=False)),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 400
|
||||
assert "credit system is disabled" in exc_info.value.detail.lower()
|
||||
|
||||
async def test_weekly_limit_exhausted_returns_400(self):
|
||||
"""When the weekly limit is also exhausted, resetting daily won't help."""
|
||||
|
||||
cfg = _make_config()
|
||||
weekly_exhausted = CoPilotUsageStatus(
|
||||
daily=UsageWindow(
|
||||
used=3_000_000,
|
||||
limit=2_500_000,
|
||||
resets_at=datetime.now(UTC) + timedelta(hours=6),
|
||||
),
|
||||
weekly=UsageWindow(
|
||||
used=12_500_000,
|
||||
limit=12_500_000,
|
||||
resets_at=datetime.now(UTC) + timedelta(days=3),
|
||||
),
|
||||
)
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()) as mock_release,
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(return_value=weekly_exhausted),
|
||||
),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 400
|
||||
assert "weekly" in exc_info.value.detail.lower()
|
||||
mock_release.assert_awaited_once()
|
||||
|
||||
async def test_redis_failure_for_reset_count_returns_503(self):
|
||||
"""When Redis is unavailable for get_daily_reset_count, returns 503."""
|
||||
|
||||
with (
|
||||
patch(f"{_MODULE}.config", _make_config()),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=None)),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 503
|
||||
assert "verify" in exc_info.value.detail.lower()
|
||||
|
||||
async def test_redis_reset_failure_refunds_credits(self):
|
||||
"""When reset_daily_usage fails, credits are refunded and 503 returned."""
|
||||
|
||||
mock_credit_model = AsyncMock()
|
||||
mock_credit_model.spend_credits.return_value = 1500
|
||||
|
||||
cfg = _make_config()
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()),
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(return_value=_usage()),
|
||||
),
|
||||
patch(
|
||||
f"{_MODULE}.get_user_credit_model",
|
||||
AsyncMock(return_value=mock_credit_model),
|
||||
),
|
||||
patch(f"{_MODULE}.reset_daily_usage", AsyncMock(return_value=False)),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 503
|
||||
assert "not been charged" in exc_info.value.detail
|
||||
mock_credit_model.top_up_credits.assert_awaited_once()
|
||||
|
||||
async def test_redis_reset_failure_refund_also_fails(self):
|
||||
"""When both reset and refund fail, error message reflects the truth."""
|
||||
|
||||
mock_credit_model = AsyncMock()
|
||||
mock_credit_model.spend_credits.return_value = 1500
|
||||
mock_credit_model.top_up_credits.side_effect = RuntimeError("db down")
|
||||
|
||||
cfg = _make_config()
|
||||
with (
|
||||
patch(f"{_MODULE}.config", cfg),
|
||||
patch(f"{_MODULE}.settings", _mock_settings()),
|
||||
_mock_rate_limits(),
|
||||
patch(f"{_MODULE}.get_daily_reset_count", AsyncMock(return_value=0)),
|
||||
patch(f"{_MODULE}.acquire_reset_lock", AsyncMock(return_value=True)),
|
||||
patch(f"{_MODULE}.release_reset_lock", AsyncMock()),
|
||||
patch(
|
||||
f"{_MODULE}.get_usage_status",
|
||||
AsyncMock(return_value=_usage()),
|
||||
),
|
||||
patch(
|
||||
f"{_MODULE}.get_user_credit_model",
|
||||
AsyncMock(return_value=mock_credit_model),
|
||||
),
|
||||
patch(f"{_MODULE}.reset_daily_usage", AsyncMock(return_value=False)),
|
||||
):
|
||||
with pytest.raises(HTTPException) as exc_info:
|
||||
await reset_copilot_usage(user_id="user-1")
|
||||
assert exc_info.value.status_code == 503
|
||||
assert "contact support" in exc_info.value.detail.lower()
|
||||
@@ -3,62 +3,26 @@
|
||||
You can create, edit, and customize agents directly. You ARE the brain —
|
||||
generate the agent JSON yourself using block schemas, then validate and save.
|
||||
|
||||
### Clarifying — Before or During Building
|
||||
|
||||
Use `ask_question` whenever the user's intent is ambiguous — whether
|
||||
that's before starting or midway through the workflow. Common moments:
|
||||
|
||||
- **Before building**: output format, delivery channel, data source, or
|
||||
trigger is unspecified.
|
||||
- **During block discovery**: multiple blocks could fit and the user
|
||||
should choose.
|
||||
- **During JSON generation**: a wiring decision depends on user
|
||||
preference.
|
||||
|
||||
Steps:
|
||||
1. Call `find_block` (or another discovery tool) to learn what the
|
||||
platform actually supports for the ambiguous dimension.
|
||||
2. Call `ask_question` with a concrete question listing the discovered
|
||||
options (e.g. "The platform supports Gmail, Slack, and Google Docs —
|
||||
which should the agent use for delivery?").
|
||||
3. **Wait for the user's answer** before continuing.
|
||||
|
||||
**Skip this** when the goal already specifies all dimensions (e.g.
|
||||
"scrape prices from Amazon and email me daily").
|
||||
|
||||
### Workflow for Creating/Editing Agents
|
||||
|
||||
1. **If editing**: First narrow to the specific agent by UUID, then fetch its
|
||||
graph: `find_library_agent(query="<agent_id>", include_graph=true)`. This
|
||||
returns the full graph structure (nodes + links). **Never edit blindly** —
|
||||
always inspect the current graph first so you know exactly what to change.
|
||||
Avoid using `include_graph=true` with broad keyword searches, as fetching
|
||||
multiple graphs at once is expensive and consumes LLM context budget.
|
||||
2. **Discover blocks**: Call `find_block(query, include_schemas=true)` to
|
||||
1. **Discover blocks**: Call `find_block(query, include_schemas=true)` to
|
||||
search for relevant blocks. This returns block IDs, names, descriptions,
|
||||
and full input/output schemas.
|
||||
3. **Find library agents**: Call `find_library_agent` to discover reusable
|
||||
2. **Find library agents**: Call `find_library_agent` to discover reusable
|
||||
agents that can be composed as sub-agents via `AgentExecutorBlock`.
|
||||
4. **Generate/modify JSON**: Build or modify the agent JSON using block schemas:
|
||||
- Use block IDs from step 2 as `block_id` in nodes
|
||||
3. **Generate JSON**: Build the agent JSON using block schemas:
|
||||
- Use block IDs from step 1 as `block_id` in nodes
|
||||
- Wire outputs to inputs using links
|
||||
- Set design-time config in `input_default`
|
||||
- Use `AgentInputBlock` for values the user provides at runtime
|
||||
- When editing, apply targeted changes and preserve unchanged parts
|
||||
5. **Write to workspace**: Save the JSON to a workspace file so the user
|
||||
4. **Write to workspace**: Save the JSON to a workspace file so the user
|
||||
can review it: `write_workspace_file(filename="agent.json", content=...)`
|
||||
6. **Validate**: Call `validate_agent_graph` with the agent JSON to check
|
||||
5. **Validate**: Call `validate_agent_graph` with the agent JSON to check
|
||||
for errors
|
||||
7. **Fix if needed**: Call `fix_agent_graph` to auto-fix common issues,
|
||||
6. **Fix if needed**: Call `fix_agent_graph` to auto-fix common issues,
|
||||
or fix manually based on the error descriptions. Iterate until valid.
|
||||
8. **Save**: Call `create_agent` (new) or `edit_agent` (existing) with
|
||||
7. **Save**: Call `create_agent` (new) or `edit_agent` (existing) with
|
||||
the final `agent_json`
|
||||
8. **Dry-run**: ALWAYS call `run_agent` with `dry_run=True` and
|
||||
`wait_for_result=120` to verify the agent works end-to-end.
|
||||
9. **Inspect & fix**: Check the dry-run output for errors. If issues are
|
||||
found, call `edit_agent` to fix and dry-run again. Repeat until the
|
||||
simulation passes or the problems are clearly unfixable.
|
||||
See "REQUIRED: Dry-Run Verification Loop" section below for details.
|
||||
|
||||
### Agent JSON Structure
|
||||
|
||||
@@ -103,17 +67,9 @@ These define the agent's interface — what it accepts and what it produces.
|
||||
**AgentInputBlock** (ID: `c0a8e994-ebf1-4a9c-a4d8-89d09c86741b`):
|
||||
- Defines a user-facing input field on the agent
|
||||
- Required `input_default` fields: `name` (str), `value` (default: null)
|
||||
- Optional: `title`, `description`
|
||||
- Optional: `title`, `description`, `placeholder_values` (for dropdowns)
|
||||
- Output: `result` — the user-provided value at runtime
|
||||
- Create one AgentInputBlock per distinct input the agent needs
|
||||
- For dropdown/select inputs, use **AgentDropdownInputBlock** instead (see below)
|
||||
|
||||
**AgentDropdownInputBlock** (ID: `655d6fdf-a334-421c-b733-520549c07cd1`):
|
||||
- Specialized input block that presents a dropdown/select to the user
|
||||
- Required `input_default` fields: `name` (str)
|
||||
- Optional: `options` (list of dropdown values; when omitted/empty, input behaves as free-text), `title`, `description`, `value` (default selection)
|
||||
- Output: `result` — the user-selected value at runtime
|
||||
- Use this instead of AgentInputBlock when the user should pick from a fixed set of options
|
||||
|
||||
**AgentOutputBlock** (ID: `363ae599-353e-4804-937e-b2ee3cef3da4`):
|
||||
- Defines a user-facing output displayed after the agent runs
|
||||
@@ -252,62 +208,19 @@ call in a loop until the task is complete:
|
||||
Regular blocks work exactly like sub-agents as tools — wire each input
|
||||
field from `source_name: "tools"` on the Orchestrator side.
|
||||
|
||||
### REQUIRED: Dry-Run Verification Loop (create -> dry-run -> fix)
|
||||
### Testing with Dry Run
|
||||
|
||||
After creating or editing an agent, you MUST dry-run it before telling the
|
||||
user the agent is ready. NEVER skip this step.
|
||||
After saving an agent, suggest a dry run to validate wiring without consuming
|
||||
real API calls, credentials, or credits:
|
||||
|
||||
#### Step-by-step workflow
|
||||
|
||||
1. **Create/Edit**: Call `create_agent` or `edit_agent` to save the agent.
|
||||
2. **Dry-run**: Call `run_agent` with `dry_run=True`, `wait_for_result=120`,
|
||||
and realistic sample inputs that exercise every path in the agent. This
|
||||
simulates execution using an LLM for each block — no real API calls,
|
||||
credentials, or credits are consumed.
|
||||
3. **Inspect output**: Examine the dry-run result for problems. If
|
||||
`wait_for_result` returns only a summary, call
|
||||
`view_agent_output(execution_id=..., show_execution_details=True)` to
|
||||
see the full node-by-node execution trace. Look for:
|
||||
- **Errors / failed nodes** — a node raised an exception or returned an
|
||||
error status. Common causes: wrong `source_name`/`sink_name` in links,
|
||||
missing `input_default` values, or referencing a nonexistent block output.
|
||||
- **Null / empty outputs** — data did not flow through a link. Verify that
|
||||
`source_name` and `sink_name` match the block schemas exactly (case-
|
||||
sensitive, including nested `_#_` notation).
|
||||
- **Nodes that never executed** — the node was not reached. Likely a
|
||||
missing or broken link from an upstream node.
|
||||
- **Unexpected values** — data arrived but in the wrong type or
|
||||
structure. Check type compatibility between linked ports.
|
||||
4. **Fix**: If any issues are found, call `edit_agent` with the corrected
|
||||
agent JSON, then go back to step 2.
|
||||
5. **Repeat**: Continue the dry-run -> fix cycle until the simulation passes
|
||||
or the problems are clearly unfixable. If you stop making progress,
|
||||
report the remaining issues to the user and ask for guidance.
|
||||
|
||||
#### Good vs bad dry-run output
|
||||
|
||||
**Good output** (agent is ready):
|
||||
- All nodes executed successfully (no errors in the execution trace)
|
||||
- Data flows through every link with non-null, correctly-typed values
|
||||
- The final `AgentOutputBlock` contains a meaningful result
|
||||
- Status is `COMPLETED`
|
||||
|
||||
**Bad output** (needs fixing):
|
||||
- Status is `FAILED` — check the error message for the failing node
|
||||
- An output node received `null` — trace back to find the broken link
|
||||
- A node received data in the wrong format (e.g. string where list expected)
|
||||
- Nodes downstream of a failing node were skipped entirely
|
||||
|
||||
**Special block behaviour in dry-run mode:**
|
||||
- **OrchestratorBlock** and **AgentExecutorBlock** execute for real so the
|
||||
orchestrator can make LLM calls and agent executors can spawn child graphs.
|
||||
Their downstream tool blocks and child-graph blocks are still simulated.
|
||||
Note: real LLM inference calls are made (consuming API quota), even though
|
||||
platform credits are not charged. Agent-mode iterations are capped at 1 in
|
||||
dry-run to keep it fast.
|
||||
- **MCPToolBlock** is simulated using the selected tool's name and JSON Schema
|
||||
so the LLM can produce a realistic mock response without connecting to the
|
||||
MCP server.
|
||||
1. **Run**: Call `run_agent` or `run_block` with `dry_run=True` and provide
|
||||
sample inputs. This executes the graph with mock outputs, verifying that
|
||||
links resolve correctly and required inputs are satisfied.
|
||||
2. **Check results**: Call `view_agent_output` with `show_execution_details=True`
|
||||
to inspect the full node-by-node execution trace. This shows what each node
|
||||
received as input and produced as output, making it easy to spot wiring issues.
|
||||
3. **Iterate**: If the dry run reveals wiring issues or missing inputs, fix
|
||||
the agent JSON and re-save before suggesting a real execution.
|
||||
|
||||
### Example: Simple AI Text Processor
|
||||
|
||||
|
||||
@@ -25,7 +25,7 @@ from backend.copilot.sdk.compaction import (
|
||||
|
||||
|
||||
def _make_session() -> ChatSession:
|
||||
return ChatSession.new(user_id="test-user", dry_run=False)
|
||||
return ChatSession.new(user_id="test-user")
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
@@ -2,30 +2,14 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from collections.abc import AsyncIterator
|
||||
from unittest.mock import patch
|
||||
from uuid import uuid4
|
||||
|
||||
import pytest
|
||||
import pytest_asyncio
|
||||
|
||||
from backend.util import json
|
||||
|
||||
|
||||
@pytest_asyncio.fixture(scope="session", loop_scope="session", name="server")
|
||||
async def _server_noop() -> None:
|
||||
"""No-op server stub — SDK tests don't need the full backend."""
|
||||
return None
|
||||
|
||||
|
||||
@pytest_asyncio.fixture(
|
||||
scope="session", loop_scope="session", autouse=True, name="graph_cleanup"
|
||||
)
|
||||
async def _graph_cleanup_noop() -> AsyncIterator[None]:
|
||||
"""No-op graph cleanup stub."""
|
||||
yield
|
||||
|
||||
|
||||
@pytest.fixture()
|
||||
def mock_chat_config():
|
||||
"""Mock ChatConfig so compact_transcript tests skip real config lookup."""
|
||||
@@ -41,64 +25,24 @@ def build_test_transcript(pairs: list[tuple[str, str]]) -> str:
|
||||
|
||||
Use this helper in any copilot SDK test that needs a well-formed
|
||||
transcript without hitting the real storage layer.
|
||||
|
||||
Delegates to ``build_structured_transcript`` — plain content strings
|
||||
are automatically wrapped in ``[{"type": "text", "text": ...}]`` for
|
||||
assistant messages.
|
||||
"""
|
||||
# Cast widening: tuple[str, str] is structurally compatible with
|
||||
# tuple[str, str | list[dict]] but list invariance requires explicit
|
||||
# annotation.
|
||||
widened: list[tuple[str, str | list[dict]]] = list(pairs)
|
||||
return build_structured_transcript(widened)
|
||||
|
||||
|
||||
def build_structured_transcript(
|
||||
entries: list[tuple[str, str | list[dict]]],
|
||||
) -> str:
|
||||
"""Build a JSONL transcript with structured content blocks.
|
||||
|
||||
Each entry is (role, content) where content is either a plain string
|
||||
(for user messages) or a list of content block dicts (for assistant
|
||||
messages with thinking/tool_use/text blocks).
|
||||
|
||||
Example::
|
||||
|
||||
build_structured_transcript([
|
||||
("user", "Hello"),
|
||||
("assistant", [
|
||||
{"type": "thinking", "thinking": "...", "signature": "sig1"},
|
||||
{"type": "text", "text": "Hi there"},
|
||||
]),
|
||||
])
|
||||
"""
|
||||
lines: list[str] = []
|
||||
last_uuid: str | None = None
|
||||
for role, content in entries:
|
||||
for role, content in pairs:
|
||||
uid = str(uuid4())
|
||||
entry_type = "assistant" if role == "assistant" else "user"
|
||||
if role == "assistant" and isinstance(content, list):
|
||||
msg: dict = {
|
||||
"role": "assistant",
|
||||
"model": "claude-test",
|
||||
"id": f"msg_{uid[:8]}",
|
||||
"type": "message",
|
||||
"content": content,
|
||||
"stop_reason": "end_turn",
|
||||
"stop_sequence": None,
|
||||
}
|
||||
elif role == "assistant":
|
||||
msg = {
|
||||
"role": "assistant",
|
||||
"model": "claude-test",
|
||||
"id": f"msg_{uid[:8]}",
|
||||
"type": "message",
|
||||
"content": [{"type": "text", "text": content}],
|
||||
"stop_reason": "end_turn",
|
||||
"stop_sequence": None,
|
||||
}
|
||||
else:
|
||||
msg = {"role": role, "content": content}
|
||||
msg: dict = {"role": role, "content": content}
|
||||
if role == "assistant":
|
||||
msg.update(
|
||||
{
|
||||
"model": "",
|
||||
"id": f"msg_{uid[:8]}",
|
||||
"type": "message",
|
||||
"content": [{"type": "text", "text": content}],
|
||||
"stop_reason": "end_turn",
|
||||
"stop_sequence": None,
|
||||
}
|
||||
)
|
||||
entry = {
|
||||
"type": entry_type,
|
||||
"uuid": uid,
|
||||
|
||||
@@ -8,9 +8,6 @@ SDK-internal paths (``~/.claude/projects/…/tool-results/``) are handled
|
||||
by the separate ``Read`` MCP tool registered in ``tool_adapter.py``.
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
import base64
|
||||
import hashlib
|
||||
import itertools
|
||||
import json
|
||||
import logging
|
||||
@@ -31,12 +28,6 @@ from backend.copilot.context import (
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# Default number of lines returned by ``read_file`` when the caller does not
|
||||
# specify a limit. Also used as the threshold in ``bridge_to_sandbox`` to
|
||||
# decide whether the model is requesting the full file (and thus whether the
|
||||
# bridge copy is worthwhile).
|
||||
_DEFAULT_READ_LIMIT = 2000
|
||||
|
||||
|
||||
async def _check_sandbox_symlink_escape(
|
||||
sandbox: Any,
|
||||
@@ -98,7 +89,7 @@ def _get_sandbox_and_path(
|
||||
return sandbox, remote
|
||||
|
||||
|
||||
async def _sandbox_write(sandbox: Any, path: str, content: str | bytes) -> None:
|
||||
async def _sandbox_write(sandbox: Any, path: str, content: str) -> None:
|
||||
"""Write *content* to *path* inside the sandbox.
|
||||
|
||||
The E2B filesystem API (``sandbox.files.write``) and the command API
|
||||
@@ -111,14 +102,11 @@ async def _sandbox_write(sandbox: Any, path: str, content: str | bytes) -> None:
|
||||
To work around this, writes targeting ``/tmp`` are performed via
|
||||
``tee`` through the command API, which runs as the sandbox ``user``
|
||||
and can therefore always overwrite user-owned files.
|
||||
|
||||
*content* may be ``str`` (text) or ``bytes`` (binary). Both paths
|
||||
are handled correctly: text is encoded to bytes for the base64 shell
|
||||
pipe, and raw bytes are passed through without any encoding.
|
||||
"""
|
||||
if path == "/tmp" or path.startswith("/tmp/"):
|
||||
raw = content.encode() if isinstance(content, str) else content
|
||||
encoded = base64.b64encode(raw).decode()
|
||||
import base64 as _b64
|
||||
|
||||
encoded = _b64.b64encode(content.encode()).decode()
|
||||
result = await sandbox.commands.run(
|
||||
f"echo {shlex.quote(encoded)} | base64 -d > {shlex.quote(path)}",
|
||||
cwd=E2B_WORKDIR,
|
||||
@@ -140,25 +128,14 @@ async def _handle_read_file(args: dict[str, Any]) -> dict[str, Any]:
|
||||
"""Read lines from a sandbox file, falling back to the local host for SDK-internal paths."""
|
||||
file_path: str = args.get("file_path", "")
|
||||
offset: int = max(0, int(args.get("offset", 0)))
|
||||
limit: int = max(1, int(args.get("limit", _DEFAULT_READ_LIMIT)))
|
||||
limit: int = max(1, int(args.get("limit", 2000)))
|
||||
|
||||
if not file_path:
|
||||
return _mcp("file_path is required", error=True)
|
||||
|
||||
# SDK-internal paths (tool-results/tool-outputs, ephemeral working dir)
|
||||
# stay on the host. When E2B is active, also copy the file into the
|
||||
# sandbox so bash_exec can access it for further processing.
|
||||
# SDK-internal paths (tool-results, ephemeral working dir) stay on the host.
|
||||
if _is_allowed_local(file_path):
|
||||
result = _read_local(file_path, offset, limit)
|
||||
if not result.get("isError"):
|
||||
sandbox = _get_sandbox()
|
||||
if sandbox is not None:
|
||||
annotation = await bridge_and_annotate(
|
||||
sandbox, file_path, offset, limit
|
||||
)
|
||||
if annotation:
|
||||
result["content"][0]["text"] += annotation
|
||||
return result
|
||||
return _read_local(file_path, offset, limit)
|
||||
|
||||
result = _get_sandbox_and_path(file_path)
|
||||
if isinstance(result, dict):
|
||||
@@ -325,103 +302,6 @@ async def _handle_grep(args: dict[str, Any]) -> dict[str, Any]:
|
||||
return _mcp(output if output else "No matches found.")
|
||||
|
||||
|
||||
# Bridging: copy SDK-internal files into E2B sandbox
|
||||
|
||||
# Files larger than this are written to /home/user/ via sandbox.files.write()
|
||||
# instead of /tmp/ via shell base64, to avoid shell argument length limits
|
||||
# and E2B command timeouts. Base64 expands content by ~33%, so keep this
|
||||
# well under the typical Linux ARG_MAX (128 KB).
|
||||
_BRIDGE_SHELL_MAX_BYTES = 32 * 1024 # 32 KB
|
||||
# Files larger than this are skipped entirely to avoid excessive transfer times.
|
||||
_BRIDGE_SKIP_BYTES = 50 * 1024 * 1024 # 50 MB
|
||||
|
||||
|
||||
async def bridge_to_sandbox(
|
||||
sandbox: Any, file_path: str, offset: int, limit: int
|
||||
) -> str | None:
|
||||
"""Best-effort copy of a host-side SDK file into the E2B sandbox.
|
||||
|
||||
When the model reads an SDK-internal file (e.g. tool-results), it often
|
||||
wants to process the data with bash. Copying the file into the sandbox
|
||||
under a stable name lets ``bash_exec`` access it without extra steps.
|
||||
|
||||
Only copies when offset=0 and limit is large enough to indicate the model
|
||||
wants the full file. Errors are logged but never propagated.
|
||||
|
||||
Returns the sandbox path on success, or ``None`` on skip/failure.
|
||||
|
||||
Size handling:
|
||||
- <= 32 KB: written to ``/tmp/<hash>-<basename>`` via shell base64
|
||||
(``_sandbox_write``). Kept small to stay within ARG_MAX.
|
||||
- 32 KB - 50 MB: written to ``/home/user/<hash>-<basename>`` via
|
||||
``sandbox.files.write()`` to avoid shell argument length limits.
|
||||
- > 50 MB: skipped entirely with a warning.
|
||||
|
||||
The sandbox filename is prefixed with a short hash of the full source
|
||||
path to avoid collisions when different source files share the same
|
||||
basename (e.g. multiple ``result.json`` files).
|
||||
"""
|
||||
if offset != 0 or limit < _DEFAULT_READ_LIMIT:
|
||||
return None
|
||||
try:
|
||||
expanded = os.path.realpath(os.path.expanduser(file_path))
|
||||
basename = os.path.basename(expanded)
|
||||
source_id = hashlib.sha256(expanded.encode()).hexdigest()[:12]
|
||||
unique_name = f"{source_id}-{basename}"
|
||||
file_size = os.path.getsize(expanded)
|
||||
if file_size > _BRIDGE_SKIP_BYTES:
|
||||
logger.warning(
|
||||
"[E2B] Skipping bridge for large file (%d bytes): %s",
|
||||
file_size,
|
||||
basename,
|
||||
)
|
||||
return None
|
||||
|
||||
def _read_bytes() -> bytes:
|
||||
with open(expanded, "rb") as fh:
|
||||
return fh.read()
|
||||
|
||||
raw_content = await asyncio.to_thread(_read_bytes)
|
||||
try:
|
||||
text_content: str | None = raw_content.decode("utf-8")
|
||||
except UnicodeDecodeError:
|
||||
text_content = None
|
||||
data: str | bytes = text_content if text_content is not None else raw_content
|
||||
if file_size <= _BRIDGE_SHELL_MAX_BYTES:
|
||||
sandbox_path = f"/tmp/{unique_name}"
|
||||
await _sandbox_write(sandbox, sandbox_path, data)
|
||||
else:
|
||||
sandbox_path = f"/home/user/{unique_name}"
|
||||
await sandbox.files.write(sandbox_path, data)
|
||||
logger.info(
|
||||
"[E2B] Bridged SDK file to sandbox: %s -> %s", basename, sandbox_path
|
||||
)
|
||||
return sandbox_path
|
||||
except Exception:
|
||||
logger.warning(
|
||||
"[E2B] Failed to bridge SDK file to sandbox: %s",
|
||||
file_path,
|
||||
exc_info=True,
|
||||
)
|
||||
return None
|
||||
|
||||
|
||||
async def bridge_and_annotate(
|
||||
sandbox: Any, file_path: str, offset: int, limit: int
|
||||
) -> str | None:
|
||||
"""Bridge a host file to the sandbox and return a newline-prefixed annotation.
|
||||
|
||||
Combines ``bridge_to_sandbox`` with the standard annotation suffix so
|
||||
callers don't need to duplicate the pattern. Returns a string like
|
||||
``"\\n[Sandbox copy available at /tmp/abc-file.txt]"`` on success, or
|
||||
``None`` if bridging was skipped or failed.
|
||||
"""
|
||||
sandbox_path = await bridge_to_sandbox(sandbox, file_path, offset, limit)
|
||||
if sandbox_path is None:
|
||||
return None
|
||||
return f"\n[Sandbox copy available at {sandbox_path}]"
|
||||
|
||||
|
||||
# Local read (for SDK-internal paths)
|
||||
|
||||
|
||||
|
||||
@@ -3,7 +3,6 @@
|
||||
Pure unit tests with no external dependencies (no E2B, no sandbox).
|
||||
"""
|
||||
|
||||
import hashlib
|
||||
import os
|
||||
import shutil
|
||||
from types import SimpleNamespace
|
||||
@@ -14,26 +13,12 @@ import pytest
|
||||
from backend.copilot.context import E2B_WORKDIR, SDK_PROJECTS_DIR, _current_project_dir
|
||||
|
||||
from .e2b_file_tools import (
|
||||
_BRIDGE_SHELL_MAX_BYTES,
|
||||
_BRIDGE_SKIP_BYTES,
|
||||
_DEFAULT_READ_LIMIT,
|
||||
_check_sandbox_symlink_escape,
|
||||
_read_local,
|
||||
_sandbox_write,
|
||||
bridge_and_annotate,
|
||||
bridge_to_sandbox,
|
||||
resolve_sandbox_path,
|
||||
)
|
||||
|
||||
|
||||
def _expected_bridge_path(file_path: str, prefix: str = "/tmp") -> str:
|
||||
"""Compute the expected sandbox path for a bridged file."""
|
||||
expanded = os.path.realpath(os.path.expanduser(file_path))
|
||||
basename = os.path.basename(expanded)
|
||||
source_id = hashlib.sha256(expanded.encode()).hexdigest()[:12]
|
||||
return f"{prefix}/{source_id}-{basename}"
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# resolve_sandbox_path — sandbox path normalisation & boundary enforcement
|
||||
# ---------------------------------------------------------------------------
|
||||
@@ -106,9 +91,9 @@ class TestResolveSandboxPath:
|
||||
# ---------------------------------------------------------------------------
|
||||
# _read_local — host filesystem reads with allowlist enforcement
|
||||
#
|
||||
# In E2B mode, _read_local only allows tool-results/tool-outputs paths
|
||||
# (via is_allowed_local_path without sdk_cwd). Regular files live on
|
||||
# the sandbox, not the host.
|
||||
# In E2B mode, _read_local only allows tool-results paths (via
|
||||
# is_allowed_local_path without sdk_cwd). Regular files live on the
|
||||
# sandbox, not the host.
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
@@ -134,7 +119,7 @@ class TestReadLocal:
|
||||
)
|
||||
token = _current_project_dir.set(encoded)
|
||||
try:
|
||||
result = _read_local(filepath, offset=0, limit=_DEFAULT_READ_LIMIT)
|
||||
result = _read_local(filepath, offset=0, limit=2000)
|
||||
assert result["isError"] is False
|
||||
assert "line 1" in result["content"][0]["text"]
|
||||
assert "line 2" in result["content"][0]["text"]
|
||||
@@ -142,25 +127,6 @@ class TestReadLocal:
|
||||
_current_project_dir.reset(token)
|
||||
os.unlink(filepath)
|
||||
|
||||
def test_read_tool_outputs_file(self):
|
||||
"""Reading a tool-outputs file should also succeed."""
|
||||
encoded = "-tmp-copilot-e2b-test-read-outputs"
|
||||
tool_outputs_dir = os.path.join(
|
||||
SDK_PROJECTS_DIR, encoded, self._CONV_UUID, "tool-outputs"
|
||||
)
|
||||
os.makedirs(tool_outputs_dir, exist_ok=True)
|
||||
filepath = os.path.join(tool_outputs_dir, "sdk-abc123.json")
|
||||
with open(filepath, "w") as f:
|
||||
f.write('{"data": "test"}\n')
|
||||
token = _current_project_dir.set(encoded)
|
||||
try:
|
||||
result = _read_local(filepath, offset=0, limit=_DEFAULT_READ_LIMIT)
|
||||
assert result["isError"] is False
|
||||
assert "test" in result["content"][0]["text"]
|
||||
finally:
|
||||
_current_project_dir.reset(token)
|
||||
shutil.rmtree(os.path.join(SDK_PROJECTS_DIR, encoded), ignore_errors=True)
|
||||
|
||||
def test_read_disallowed_path_blocked(self):
|
||||
"""Reading /etc/passwd should be blocked by the allowlist."""
|
||||
result = _read_local("/etc/passwd", offset=0, limit=10)
|
||||
@@ -369,199 +335,3 @@ class TestSandboxWrite:
|
||||
encoded_in_cmd = call_args.split("echo ")[1].split(" |")[0].strip("'")
|
||||
decoded = base64.b64decode(encoded_in_cmd).decode()
|
||||
assert decoded == content
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# bridge_to_sandbox — copy SDK-internal files into E2B sandbox
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
def _make_bridge_sandbox() -> SimpleNamespace:
|
||||
"""Build a sandbox mock suitable for bridge_to_sandbox tests."""
|
||||
run_result = SimpleNamespace(stdout="", stderr="", exit_code=0)
|
||||
commands = SimpleNamespace(run=AsyncMock(return_value=run_result))
|
||||
files = SimpleNamespace(write=AsyncMock())
|
||||
return SimpleNamespace(commands=commands, files=files)
|
||||
|
||||
|
||||
class TestBridgeToSandbox:
|
||||
@pytest.mark.asyncio
|
||||
async def test_happy_path_small_file(self, tmp_path):
|
||||
"""A small file is bridged to /tmp/<hash>-<basename> via _sandbox_write."""
|
||||
f = tmp_path / "result.json"
|
||||
f.write_text('{"ok": true}')
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
expected = _expected_bridge_path(str(f))
|
||||
assert result == expected
|
||||
sandbox.commands.run.assert_called_once()
|
||||
cmd = sandbox.commands.run.call_args[0][0]
|
||||
assert "result.json" in cmd
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skip_when_offset_nonzero(self, tmp_path):
|
||||
"""Bridging is skipped when offset != 0 (partial read)."""
|
||||
f = tmp_path / "data.txt"
|
||||
f.write_text("content")
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=10, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
assert result is None
|
||||
sandbox.commands.run.assert_not_called()
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skip_when_limit_too_small(self, tmp_path):
|
||||
"""Bridging is skipped when limit < _DEFAULT_READ_LIMIT (partial read)."""
|
||||
f = tmp_path / "data.txt"
|
||||
f.write_text("content")
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
await bridge_to_sandbox(sandbox, str(f), offset=0, limit=100)
|
||||
|
||||
sandbox.commands.run.assert_not_called()
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_nonexistent_file_does_not_raise(self, tmp_path):
|
||||
"""Bridging a non-existent file logs but does not propagate errors."""
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
await bridge_to_sandbox(
|
||||
sandbox, str(tmp_path / "ghost.txt"), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
sandbox.commands.run.assert_not_called()
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_sandbox_write_failure_returns_none(self, tmp_path):
|
||||
"""If sandbox write fails, returns None (best-effort)."""
|
||||
f = tmp_path / "data.txt"
|
||||
f.write_text("content")
|
||||
sandbox = _make_bridge_sandbox()
|
||||
sandbox.commands.run.side_effect = RuntimeError("E2B timeout")
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
assert result is None
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_large_file_uses_files_api(self, tmp_path):
|
||||
"""Files > 32 KB but <= 50 MB are written to /home/user/ via files.write."""
|
||||
f = tmp_path / "big.json"
|
||||
f.write_bytes(b"x" * (_BRIDGE_SHELL_MAX_BYTES + 1))
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
expected = _expected_bridge_path(str(f), prefix="/home/user")
|
||||
assert result == expected
|
||||
sandbox.files.write.assert_called_once()
|
||||
call_args = sandbox.files.write.call_args[0]
|
||||
assert call_args[0] == expected
|
||||
sandbox.commands.run.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_small_binary_file_preserves_bytes(self, tmp_path):
|
||||
"""A small binary file is bridged to /tmp via base64 without corruption."""
|
||||
binary_data = bytes(range(256))
|
||||
f = tmp_path / "image.png"
|
||||
f.write_bytes(binary_data)
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
expected = _expected_bridge_path(str(f))
|
||||
assert result == expected
|
||||
sandbox.commands.run.assert_called_once()
|
||||
cmd = sandbox.commands.run.call_args[0][0]
|
||||
assert "base64" in cmd
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_large_binary_file_writes_raw_bytes(self, tmp_path):
|
||||
"""A large binary file is bridged to /home/user/ as raw bytes."""
|
||||
binary_data = bytes(range(256)) * 200
|
||||
f = tmp_path / "photo.jpg"
|
||||
f.write_bytes(binary_data)
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
expected = _expected_bridge_path(str(f), prefix="/home/user")
|
||||
assert result == expected
|
||||
sandbox.files.write.assert_called_once()
|
||||
call_args = sandbox.files.write.call_args[0]
|
||||
assert call_args[0] == expected
|
||||
assert call_args[1] == binary_data
|
||||
sandbox.commands.run.assert_not_called()
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_very_large_file_skipped(self, tmp_path):
|
||||
"""Files > 50 MB are skipped entirely."""
|
||||
f = tmp_path / "huge.bin"
|
||||
# Create a sparse file to avoid actually writing 50 MB
|
||||
with open(f, "wb") as fh:
|
||||
fh.seek(_BRIDGE_SKIP_BYTES + 1)
|
||||
fh.write(b"\0")
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
result = await bridge_to_sandbox(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
assert result is None
|
||||
|
||||
sandbox.commands.run.assert_not_called()
|
||||
sandbox.files.write.assert_not_called()
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# bridge_and_annotate — shared helper wrapping bridge_to_sandbox + annotation
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestBridgeAndAnnotate:
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_annotation_on_success(self, tmp_path):
|
||||
"""On success, returns a newline-prefixed annotation with the sandbox path."""
|
||||
f = tmp_path / "data.json"
|
||||
f.write_text('{"ok": true}')
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
annotation = await bridge_and_annotate(
|
||||
sandbox, str(f), offset=0, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
expected_path = _expected_bridge_path(str(f))
|
||||
assert annotation == f"\n[Sandbox copy available at {expected_path}]"
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_returns_none_when_skipped(self, tmp_path):
|
||||
"""When bridging is skipped (e.g. offset != 0), returns None."""
|
||||
f = tmp_path / "data.json"
|
||||
f.write_text("content")
|
||||
sandbox = _make_bridge_sandbox()
|
||||
|
||||
annotation = await bridge_and_annotate(
|
||||
sandbox, str(f), offset=10, limit=_DEFAULT_READ_LIMIT
|
||||
)
|
||||
|
||||
assert annotation is None
|
||||
|
||||
@@ -275,7 +275,7 @@ class TestCompactionE2E:
|
||||
|
||||
# --- Step 7: CompactionTracker receives PreCompact hook ---
|
||||
tracker = CompactionTracker()
|
||||
session = ChatSession.new(user_id="test-user", dry_run=False)
|
||||
session = ChatSession.new(user_id="test-user")
|
||||
tracker.on_compact(str(session_file))
|
||||
|
||||
# --- Step 8: Next SDK message arrives → emit_start ---
|
||||
@@ -376,7 +376,7 @@ class TestCompactionE2E:
|
||||
monkeypatch.setenv("CLAUDE_CONFIG_DIR", str(config_dir))
|
||||
|
||||
tracker = CompactionTracker()
|
||||
session = ChatSession.new(user_id="test", dry_run=False)
|
||||
session = ChatSession.new(user_id="test")
|
||||
builder = TranscriptBuilder()
|
||||
|
||||
# --- First query with compaction ---
|
||||
|
||||
@@ -1,82 +0,0 @@
|
||||
"""SDK environment variable builder — importable without circular deps.
|
||||
|
||||
Extracted from ``service.py`` so that ``backend.blocks.orchestrator``
|
||||
can reuse the same subscription / OpenRouter / direct-Anthropic logic
|
||||
without pulling in the full copilot service module (which would create a
|
||||
circular import through ``executor`` → ``credit`` → ``block_cost_config``).
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from backend.copilot.config import ChatConfig
|
||||
from backend.copilot.sdk.subscription import validate_subscription
|
||||
|
||||
# ChatConfig is stateless (reads env vars) — a separate instance is fine.
|
||||
# A singleton would require importing service.py which causes the circular dep
|
||||
# this module was created to avoid.
|
||||
config = ChatConfig()
|
||||
|
||||
|
||||
def build_sdk_env(
|
||||
session_id: str | None = None,
|
||||
user_id: str | None = None,
|
||||
sdk_cwd: str | None = None,
|
||||
) -> dict[str, str]:
|
||||
"""Build env vars for the SDK CLI subprocess.
|
||||
|
||||
Three modes (checked in order):
|
||||
1. **Subscription** — clears all keys; CLI uses ``claude login`` auth.
|
||||
2. **Direct Anthropic** — returns ``{}``; subprocess inherits
|
||||
``ANTHROPIC_API_KEY`` from the parent environment.
|
||||
3. **OpenRouter** (default) — overrides base URL and auth token to
|
||||
route through the proxy, with Langfuse trace headers.
|
||||
|
||||
When *sdk_cwd* is provided, ``CLAUDE_CODE_TMPDIR`` is set so that
|
||||
the CLI writes temp/sub-agent output inside the per-session workspace
|
||||
directory rather than an inaccessible system temp path.
|
||||
"""
|
||||
# --- Mode 1: Claude Code subscription auth ---
|
||||
if config.use_claude_code_subscription:
|
||||
validate_subscription()
|
||||
env: dict[str, str] = {
|
||||
"ANTHROPIC_API_KEY": "",
|
||||
"ANTHROPIC_AUTH_TOKEN": "",
|
||||
"ANTHROPIC_BASE_URL": "",
|
||||
}
|
||||
if sdk_cwd:
|
||||
env["CLAUDE_CODE_TMPDIR"] = sdk_cwd
|
||||
return env
|
||||
|
||||
# --- Mode 2: Direct Anthropic (no proxy hop) ---
|
||||
if not config.openrouter_active:
|
||||
env = {}
|
||||
if sdk_cwd:
|
||||
env["CLAUDE_CODE_TMPDIR"] = sdk_cwd
|
||||
return env
|
||||
|
||||
# --- Mode 3: OpenRouter proxy ---
|
||||
base = (config.base_url or "").rstrip("/")
|
||||
if base.endswith("/v1"):
|
||||
base = base[:-3]
|
||||
env = {
|
||||
"ANTHROPIC_BASE_URL": base,
|
||||
"ANTHROPIC_AUTH_TOKEN": config.api_key or "",
|
||||
"ANTHROPIC_API_KEY": "", # force CLI to use AUTH_TOKEN
|
||||
}
|
||||
|
||||
# Inject broadcast headers so OpenRouter forwards traces to Langfuse.
|
||||
def _safe(v: str) -> str:
|
||||
return v.replace("\r", "").replace("\n", "").strip()[:128]
|
||||
|
||||
parts = []
|
||||
if session_id:
|
||||
parts.append(f"x-session-id: {_safe(session_id)}")
|
||||
if user_id:
|
||||
parts.append(f"x-user-id: {_safe(user_id)}")
|
||||
if parts:
|
||||
env["ANTHROPIC_CUSTOM_HEADERS"] = "\n".join(parts)
|
||||
|
||||
if sdk_cwd:
|
||||
env["CLAUDE_CODE_TMPDIR"] = sdk_cwd
|
||||
|
||||
return env
|
||||
@@ -1,293 +0,0 @@
|
||||
"""Tests for build_sdk_env() — the SDK subprocess environment builder."""
|
||||
|
||||
from unittest.mock import patch
|
||||
|
||||
import pytest
|
||||
|
||||
from backend.copilot.config import ChatConfig
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Helpers — build a ChatConfig with explicit field values so tests don't
|
||||
# depend on real environment variables.
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
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)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Mode 1 — Subscription auth
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestBuildSdkEnvSubscription:
|
||||
"""When ``use_claude_code_subscription`` is True, keys are blanked."""
|
||||
|
||||
@patch("backend.copilot.sdk.env.validate_subscription")
|
||||
def test_returns_blanked_keys(self, mock_validate):
|
||||
"""Subscription mode clears API_KEY, AUTH_TOKEN, and BASE_URL."""
|
||||
cfg = _make_config(use_claude_code_subscription=True)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result == {
|
||||
"ANTHROPIC_API_KEY": "",
|
||||
"ANTHROPIC_AUTH_TOKEN": "",
|
||||
"ANTHROPIC_BASE_URL": "",
|
||||
}
|
||||
mock_validate.assert_called_once()
|
||||
|
||||
@patch(
|
||||
"backend.copilot.sdk.env.validate_subscription",
|
||||
side_effect=RuntimeError("CLI not found"),
|
||||
)
|
||||
def test_propagates_validation_error(self, mock_validate):
|
||||
"""If validate_subscription fails, the error bubbles up."""
|
||||
cfg = _make_config(use_claude_code_subscription=True)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
with pytest.raises(RuntimeError, match="CLI not found"):
|
||||
build_sdk_env()
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Mode 2 — Direct Anthropic (no OpenRouter)
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestBuildSdkEnvDirectAnthropic:
|
||||
"""When OpenRouter is inactive, return empty dict (inherit parent env)."""
|
||||
|
||||
def test_returns_empty_dict_when_openrouter_inactive(self):
|
||||
cfg = _make_config(use_openrouter=False)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result == {}
|
||||
|
||||
def test_returns_empty_dict_when_openrouter_flag_true_but_no_key(self):
|
||||
"""OpenRouter flag is True but no api_key => openrouter_active is False."""
|
||||
cfg = _make_config(use_openrouter=True, base_url="https://openrouter.ai/api/v1")
|
||||
# Force api_key to None after construction (field_validator may pick up env vars)
|
||||
object.__setattr__(cfg, "api_key", None)
|
||||
assert not cfg.openrouter_active
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result == {}
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Mode 3 — OpenRouter proxy
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestBuildSdkEnvOpenRouter:
|
||||
"""When OpenRouter is active, return proxy env vars."""
|
||||
|
||||
def _openrouter_config(self, **overrides):
|
||||
defaults = {
|
||||
"use_openrouter": True,
|
||||
"api_key": "sk-or-test-key",
|
||||
"base_url": "https://openrouter.ai/api/v1",
|
||||
}
|
||||
defaults.update(overrides)
|
||||
return _make_config(**defaults)
|
||||
|
||||
def test_basic_openrouter_env(self):
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result["ANTHROPIC_BASE_URL"] == "https://openrouter.ai/api"
|
||||
assert result["ANTHROPIC_AUTH_TOKEN"] == "sk-or-test-key"
|
||||
assert result["ANTHROPIC_API_KEY"] == ""
|
||||
assert "ANTHROPIC_CUSTOM_HEADERS" not in result
|
||||
|
||||
def test_strips_trailing_v1(self):
|
||||
"""The /v1 suffix is stripped from the base URL."""
|
||||
cfg = self._openrouter_config(base_url="https://openrouter.ai/api/v1")
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result["ANTHROPIC_BASE_URL"] == "https://openrouter.ai/api"
|
||||
|
||||
def test_strips_trailing_v1_and_slash(self):
|
||||
"""Trailing slash before /v1 strip is handled."""
|
||||
cfg = self._openrouter_config(base_url="https://openrouter.ai/api/v1/")
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
# rstrip("/") first, then remove /v1
|
||||
assert result["ANTHROPIC_BASE_URL"] == "https://openrouter.ai/api"
|
||||
|
||||
def test_no_v1_suffix_left_alone(self):
|
||||
"""A base URL without /v1 is used as-is."""
|
||||
cfg = self._openrouter_config(base_url="https://custom-proxy.example.com")
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
assert result["ANTHROPIC_BASE_URL"] == "https://custom-proxy.example.com"
|
||||
|
||||
def test_session_id_header(self):
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(session_id="sess-123")
|
||||
|
||||
assert "ANTHROPIC_CUSTOM_HEADERS" in result
|
||||
assert "x-session-id: sess-123" in result["ANTHROPIC_CUSTOM_HEADERS"]
|
||||
|
||||
def test_user_id_header(self):
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(user_id="user-456")
|
||||
|
||||
assert "x-user-id: user-456" in result["ANTHROPIC_CUSTOM_HEADERS"]
|
||||
|
||||
def test_both_headers(self):
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(session_id="s1", user_id="u2")
|
||||
|
||||
headers = result["ANTHROPIC_CUSTOM_HEADERS"]
|
||||
assert "x-session-id: s1" in headers
|
||||
assert "x-user-id: u2" in headers
|
||||
# They should be newline-separated
|
||||
assert "\n" in headers
|
||||
|
||||
def test_header_sanitisation_strips_newlines(self):
|
||||
"""Newlines/carriage-returns in header values are stripped."""
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(session_id="bad\r\nvalue")
|
||||
|
||||
header_val = result["ANTHROPIC_CUSTOM_HEADERS"]
|
||||
# The _safe helper removes \r and \n
|
||||
assert "\r" not in header_val.split(": ", 1)[1]
|
||||
assert "badvalue" in header_val
|
||||
|
||||
def test_header_value_truncated_to_128_chars(self):
|
||||
"""Header values are truncated to 128 characters."""
|
||||
cfg = self._openrouter_config()
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
long_id = "x" * 200
|
||||
result = build_sdk_env(session_id=long_id)
|
||||
|
||||
# The value after "x-session-id: " should be at most 128 chars
|
||||
header_line = result["ANTHROPIC_CUSTOM_HEADERS"]
|
||||
value = header_line.split(": ", 1)[1]
|
||||
assert len(value) == 128
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Mode priority
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestBuildSdkEnvModePriority:
|
||||
"""Subscription mode takes precedence over OpenRouter."""
|
||||
|
||||
@patch("backend.copilot.sdk.env.validate_subscription")
|
||||
def test_subscription_overrides_openrouter(self, mock_validate):
|
||||
cfg = _make_config(
|
||||
use_claude_code_subscription=True,
|
||||
use_openrouter=True,
|
||||
api_key="sk-or-key",
|
||||
base_url="https://openrouter.ai/api/v1",
|
||||
)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env()
|
||||
|
||||
# Should get subscription result, not OpenRouter
|
||||
assert result == {
|
||||
"ANTHROPIC_API_KEY": "",
|
||||
"ANTHROPIC_AUTH_TOKEN": "",
|
||||
"ANTHROPIC_BASE_URL": "",
|
||||
}
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# CLAUDE_CODE_TMPDIR integration
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class TestClaudeCodeTmpdir:
|
||||
"""Verify build_sdk_env() sets CLAUDE_CODE_TMPDIR from *sdk_cwd*."""
|
||||
|
||||
def test_tmpdir_set_when_sdk_cwd_is_truthy(self):
|
||||
"""CLAUDE_CODE_TMPDIR is set to sdk_cwd when sdk_cwd is truthy."""
|
||||
cfg = _make_config(use_openrouter=False)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(sdk_cwd="/tmp/copilot-workspace")
|
||||
|
||||
assert result["CLAUDE_CODE_TMPDIR"] == "/tmp/copilot-workspace"
|
||||
|
||||
def test_tmpdir_not_set_when_sdk_cwd_is_none(self):
|
||||
"""CLAUDE_CODE_TMPDIR is NOT in the env when sdk_cwd is None."""
|
||||
cfg = _make_config(use_openrouter=False)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(sdk_cwd=None)
|
||||
|
||||
assert "CLAUDE_CODE_TMPDIR" not in result
|
||||
|
||||
def test_tmpdir_not_set_when_sdk_cwd_is_empty_string(self):
|
||||
"""CLAUDE_CODE_TMPDIR is NOT in the env when sdk_cwd is empty string."""
|
||||
cfg = _make_config(use_openrouter=False)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(sdk_cwd="")
|
||||
|
||||
assert "CLAUDE_CODE_TMPDIR" not in result
|
||||
|
||||
@patch("backend.copilot.sdk.env.validate_subscription")
|
||||
def test_tmpdir_set_in_subscription_mode(self, mock_validate):
|
||||
"""CLAUDE_CODE_TMPDIR is set even in subscription mode."""
|
||||
cfg = _make_config(use_claude_code_subscription=True)
|
||||
with patch("backend.copilot.sdk.env.config", cfg):
|
||||
from backend.copilot.sdk.env import build_sdk_env
|
||||
|
||||
result = build_sdk_env(sdk_cwd="/tmp/sub-workspace")
|
||||
|
||||
assert result["CLAUDE_CODE_TMPDIR"] == "/tmp/sub-workspace"
|
||||
assert result["ANTHROPIC_API_KEY"] == ""
|
||||
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user