## Summary
This PR introduces two explicit safe mode toggles for controlling agent
execution behavior, providing clearer and more granular control over
when agents should pause for human review.
### Key Changes
**New Safe Mode Settings:**
- **`human_in_the_loop_safe_mode`** (bool, default `true`) - Controls
whether human-in-the-loop (HITL) blocks pause for review
- **`sensitive_action_safe_mode`** (bool, default `false`) - Controls
whether sensitive action blocks pause for review
**New Computed Properties on LibraryAgent:**
- `has_human_in_the_loop` - Indicates if agent contains HITL blocks
- `has_sensitive_action` - Indicates if agent contains sensitive action
blocks
**Block Changes:**
- Renamed `requires_human_review` to `is_sensitive_action` on blocks for
clarity
- Blocks marked as `is_sensitive_action=True` pause only when
`sensitive_action_safe_mode=True`
- HITL blocks pause when `human_in_the_loop_safe_mode=True`
**Frontend Changes:**
- Two separate toggles in Agent Settings based on block types present
- Toggle visibility based on `has_human_in_the_loop` and
`has_sensitive_action` computed properties
- Settings cog hidden if neither toggle applies
- Proper state management for both toggles with defaults
**AI-Generated Agent Behavior:**
- AI-generated agents set `sensitive_action_safe_mode=True` by default
- This ensures sensitive actions are reviewed for AI-generated content
## Changes
**Backend:**
- `backend/data/graph.py` - Updated `GraphSettings` with two boolean
toggles (non-optional with defaults), added `has_sensitive_action`
computed property
- `backend/data/block.py` - Renamed `requires_human_review` to
`is_sensitive_action`, updated review logic
- `backend/data/execution.py` - Updated `ExecutionContext` with both
safe mode fields
- `backend/api/features/library/model.py` - Added
`has_human_in_the_loop` and `has_sensitive_action` to `LibraryAgent`
- `backend/api/features/library/db.py` - Updated to use
`sensitive_action_safe_mode` parameter
- `backend/executor/utils.py` - Simplified execution context creation
**Frontend:**
- `useAgentSafeMode.ts` - Rewritten to support two independent toggles
- `AgentSettingsModal.tsx` - Shows two separate toggles
- `SelectedSettingsView.tsx` - Shows two separate toggles
- Regenerated API types with new schema
## Test Plan
- [x] All backend tests pass (Python 3.11, 3.12, 3.13)
- [x] All frontend tests pass
- [x] Backend format and lint pass
- [x] Frontend format and lint pass
- [x] Pre-commit hooks pass
---------
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
## Summary
- Fix intermittent "type 'vector' does not exist" errors when using
PgBouncer in transaction mode
- The issue was that `SET search_path` and the actual query could run on
different backend connections
- Use explicit schema qualification (`{schema}.vector`,
`OPERATOR({schema}.<=>)`) instead of relying on search_path
## Test plan
- [x] Tested vector type cast on local: `'[1,2,3]'::platform.vector`
works
- [x] Tested OPERATOR syntax on local: `OPERATOR(platform.<=>)` works
- [x] Tested on dev via kubectl exec: both work correctly
- [ ] Deploy to dev and verify backfill_missing_embeddings endpoint no
longer errors
## Related Issues
Fixes: AUTOGPT-SERVER-763, AUTOGPT-SERVER-764
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
# feat(backend/blocks): add ConcatenateListsBlock
## Description
This PR implements a new block `ConcatenateListsBlock` that concatenates
multiple lists into a single list. This addresses the "good first issue"
for implementing a list concatenation block in the platform/blocks area.
The block takes a list of lists as input and combines all elements in
order into a single concatenated list. This is useful for workflows that
need to merge data from multiple sources or combine results from
different operations.
### Changes 🏗️
- **Added `ConcatenateListsBlock` class** in
`autogpt_platform/backend/backend/blocks/data_manipulation.py`
- Input: `lists: List[List[Any]]` - accepts a list of lists to
concatenate
- Output: `concatenated_list: List[Any]` - returns a single concatenated
list
- Error output: `error: str` - provides clear error messages for invalid
input types
- Block ID: `3cf9298b-5817-4141-9d80-7c2cc5199c8e`
- Category: `BlockCategory.BASIC` (consistent with other list
manipulation blocks)
- **Added comprehensive test suite** in
`autogpt_platform/backend/test/blocks/test_concatenate_lists.py`
- Tests using built-in `test_input`/`test_output` validation
- Manual test cases covering edge cases (empty lists, single list, empty
input)
- Error handling tests for invalid input types
- Category consistency verification
- All tests passing
- **Implementation details:**
- Uses `extend()` method for efficient list concatenation
- Preserves element order from all input lists
- **Runtime type validation**: Explicitly checks `isinstance(lst, list)`
before calling `extend()` to prevent:
- Strings being iterated character-by-character (e.g., `extend("abc")` →
`['a', 'b', 'c']`)
- Non-iterable types causing `TypeError` (e.g., `extend(1)`)
- Clear error messages indicating which index has invalid input
- Handles edge cases: empty lists, empty input, single list, None values
- Follows existing block patterns and conventions
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Run `poetry run pytest test/blocks/test_concatenate_lists.py -v` -
all tests pass
- [x] Verified block can be imported and instantiated
- [x] Tested with built-in test cases (4 test scenarios)
- [x] Tested manual edge cases (empty lists, single list, empty input)
- [x] Tested error handling for invalid input types
- [x] Verified category is `BASIC` for consistency
- [x] Verified no linting errors
- [x] Confirmed block follows same patterns as other blocks in
`data_manipulation.py`
#### Code Quality:
- [x] Code follows existing patterns and conventions
- [x] Type hints are properly used
- [x] Documentation strings are clear and descriptive
- [x] Runtime type validation implemented
- [x] Error handling with clear error messages
- [x] No linting errors
- [x] Prisma client generated successfully
### Testing
**Test Results:**
```
test/blocks/test_concatenate_lists.py::test_concatenate_lists_block_builtin_tests PASSED
test/blocks/test_concatenate_lists.py::test_concatenate_lists_manual PASSED
============================== 2 passed in 8.35s ==============================
```
**Test Coverage:**
- Basic concatenation: `[[1, 2, 3], [4, 5, 6]]` → `[1, 2, 3, 4, 5, 6]`
- Mixed types: `[["a", "b"], ["c"], ["d", "e", "f"]]` → `["a", "b", "c",
"d", "e", "f"]`
- Empty list handling: `[[1, 2], []]` → `[1, 2]`
- Empty input: `[]` → `[]`
- Single list: `[[1, 2, 3]]` → `[1, 2, 3]`
- Error handling: Invalid input types (strings, non-lists) produce clear
error messages
- Category verification: Confirmed `BlockCategory.BASIC` for consistency
### Review Feedback Addressed
- **Category Consistency**: Changed from `BlockCategory.DATA` to
`BlockCategory.BASIC` to match other list manipulation blocks
(`AddToListBlock`, `FindInListBlock`, etc.)
- **Type Robustness**: Added explicit runtime validation with
`isinstance(lst, list)` check before calling `extend()` to prevent:
- Strings being iterated character-by-character
- Non-iterable types causing `TypeError`
- **Error Handling**: Added `error` output field with clear, descriptive
error messages indicating which index has invalid input
- **Test Coverage**: Added test case for error handling with invalid
input types
### Related Issues
- Addresses: "Implement block to concatenate lists" (good first issue,
platform/blocks, hacktoberfest)
### Notes
- This is a straightforward data manipulation block that doesn't require
external dependencies
- The block will be automatically discovered by the block loading system
- No database or configuration changes required
- Compatible with existing workflow system
- All review feedback has been addressed and incorporated
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Adds a new list utility and updates docs.
>
> - **New block**: `ConcatenateListsBlock` in
`backend/blocks/data_manipulation.py`
> - Input `lists: List[List[Any]]`; outputs `concatenated_list` or
`error`
> - Skips `None` entries; emits error for non-list items; preserves
order
> - **Docs**: Adds "Concatenate Lists" section to
`docs/integrations/basic.md` and links it in
`docs/integrations/README.md`
> - **Contributor guide**: New `docs/CLAUDE.md` with manual doc section
guidelines
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
4f56dd86c2. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Zamil Majdy <zamil.majdy@agpt.co>
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Update black version to match pre-commit hook (24.10.0) and reformat
all files with the new version.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add missing strategies (lats, multi_agent_debate) to PromptStrategyName
- Fix method override signatures for reasoning_effort parameter
- Fix Pydantic Field() overload issues with helper function
- Fix BeautifulSoup Tag type narrowing in web_fetch.py
- Fix Optional member access in playwright_browser.py and rewoo.py
- Convert hasattr patterns to getattr for proper type narrowing
- Add proper type casts for Literal types
- Fix file storage path type conversions
- Exclude legacy challenges/ from pyright checking
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
## Background
When using chat to run blocks/agents that support multiple credential
types (e.g., GitHub blocks support both `api_key` and `oauth2`), users
reported that the credentials setup UI would randomly show either "Add
API key" or "Connect account (OAuth)" - seemingly at random between
requests or server restarts.
## Root Cause
The bug was in how the backend selected which credential type to return
when building the missing credentials response:
```python
cred_type = next(iter(field_info.supported_types), "api_key")
```
The problem is that `supported_types` is a **frozenset**. When you call
`iter()` on a frozenset and take `next()`, the iteration order is
**non-deterministic** due to Python's hash randomization. This means:
- `frozenset({'api_key', 'oauth2'})` could iterate as either
`['api_key', 'oauth2']` or `['oauth2', 'api_key']`
- The order varies between Python process restarts and sometimes between
requests
- This caused the UI to randomly show different credential options
### Changes 🏗️
**Backend (`utils.py`, `run_block.py`, `run_agent.py`):**
- Added `_serialize_missing_credential()` helper that uses `sorted()`
for deterministic ordering
- Added `build_missing_credentials_from_graph()` and
`build_missing_credentials_from_field_info()` utilities
- Now returns both `type` (first sorted type, for backwards compat) and
`types` (full array with ALL supported types)
**Frontend (`helpers.ts`, `ChatCredentialsSetup.tsx`,
`useChatMessage.ts`):**
- Updated to read the `types` array from backend response
- Changed `credentialType` (single) to `credentialTypes` (array)
throughout the chat credentials flow
- Passes all supported types to `CredentialsInput` via
`credentials_types` schema field
### Result
Now `useCredentials.ts` correctly sets both `supportsApiKey=true` AND
`supportsOAuth2=true` when both are supported, ensuring:
1. **Deterministic behavior** - no more random type selection
2. **All saved credentials shown** - credentials of any supported type
appear in the selection list
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Verified GitHub block shows consistent credential options across
page reloads
- [x] Verified both OAuth and API key credentials appear in selection
when user has both saved
- [x] Verified backend returns `types: ["api_key", "oauth2"]` array
(checked via Python REPL)
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Ensures deterministic credential type selection and surfaces all
supported types end-to-end.
>
> - Backend: add `_serialize_missing_credential`,
`build_missing_credentials_from_graph/field_info`;
`run_agent`/`run_block` now return missing credentials with stable
ordering and both `type` (first) and `types` (all).
> - Frontend: chat helpers and UI (`helpers.ts`,
`ChatCredentialsSetup.tsx`, `useChatMessage.ts`) now read `types`,
switch from single `credentialType` to `credentialTypes`, and pass all
supported `credentials_types` in schemas.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
7d80f4f0e0. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com>
Co-authored-by: Nicholas Tindle <ntindle@users.noreply.github.com>
### Changes 🏗️
- Enhanced UI for the Run Graph button with improved loading states and
animations
- Added color-coded edges in the flow editor based on output data types
- Improved the layout of the Run Input Dialog with a two-column grid
design
- Refined the styling of flow editor controls with consistent icon sizes
and colors
- Updated tutorial icons with better color and size customization
- Fixed credential field display to show provider name with "credential"
suffix
- Optimized draft saving by excluding node position changes to prevent
excessive saves when dragging nodes
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Verified that the Run Graph button shows proper loading states
- [x] Confirmed that edges display correct colors based on data types
- [x] Tested the Run Input Dialog layout with various input
configurations
- [x] Checked that flow editor controls display consistently
- [x] Verified that tutorial icons render properly
- [x] Confirmed credential fields show proper provider names
- [x] Tested that dragging nodes doesn't trigger unnecessary draft saves
### Changes 🏗️
- Refactored the credentials input handling in the RunInputDialog to use
the shared CredentialsGroupedView component
- Moved CredentialsGroupedView from agent library to a shared component
location for reuse
- Fixed source name handling in edge creation to properly handle tool
source names
- Improved node output UI by replacing custom expand/collapse with
Accordion component
- Fixed timing of hardcoded values synchronization with handle IDs to
ensure proper loading
- Enabled NEW_FLOW_EDITOR and BUILDER_VIEW_SWITCH feature flags by
default
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Verified credentials input works in both agent run dialog and
builder run dialog
- [x] Confirmed node output accordion works correctly
- [x] Tested flow editor with tools to ensure source name handling works
properly
- [x] Verified hardcoded values sync correctly with handle IDs
#### For configuration changes:
- [x] `.env.default` is updated or already compatible with my changes
- [x] `docker-compose.yml` is updated or already compatible with my
changes
- [x] I have included a list of my configuration changes in the PR
description (under **Changes**)
- Consolidate classic poetry-install hooks into single hook using classic/
- Update isort hook to work with consolidated project structure
- Simplify flake8 hooks to use single classic/.flake8 config
- Consolidate pyright hooks into single hook for classic/
- Add direct_benchmark to hook coverage
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Update .flake8 config to exclude workspace directories and ignore E203
- Fix import sorting (isort) across multiple files
- Fix code formatting (black) across multiple files
- Remove unused imports and fix line length issues (flake8)
- Fix f-strings without placeholders and unused variables
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add comprehensive sub-agent spawning infrastructure that enables prompt
strategies to coordinate multiple agents for advanced reasoning patterns.
New files:
- forge/agent/execution_context.py: ExecutionContext, ResourceBudget,
SubAgentHandle, and AgentFactory protocol for sub-agent lifecycle
- agent_factory/default_factory.py: DefaultAgentFactory implementation
- prompt_strategies/lats.py: Language Agent Tree Search using MCTS
with sub-agents for action expansion and evaluation
- prompt_strategies/multi_agent_debate.py: Multi-agent debate with
proposal, critique, and consensus phases
Key changes:
- BaseMultiStepPromptStrategy gains spawn_sub_agent(), run_sub_agent(),
spawn_and_run(), and run_parallel() methods
- Agent class accepts optional ExecutionContext and injects it into strategies
- Sub-agents enabled by default (enable_sub_agents=True)
- Resource limits: max_depth=5, max_sub_agents=25, max_cycles=25
All 7 strategies now available in benchmark:
one_shot, rewoo, plan_execute, reflexion, tree_of_thoughts, lats, multi_agent_debate
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Update all classic CI workflows to use the single consolidated
pyproject.toml at classic/ instead of individual project directories.
Changes:
- classic-autogpt-ci.yml: Run from classic/, update cache key and test paths
- classic-forge-ci.yml: Run from classic/, update cache key and test paths
- classic-benchmark-ci.yml: Run from classic/, use direct-benchmark command
- classic-python-checks.yml: Simplify to single job (no matrix needed)
- classic-autogpts-ci.yml: Update to use direct-benchmark for smoke tests
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Enable agents to execute shell commands during benchmarks by setting
execute_local_commands=True and using denylist mode to block dangerous
commands (rm, sudo, chmod, kill, etc.) while allowing safe operations.
Also adds ExecutePython challenge to test code execution capability.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Merge forge/, original_autogpt/, and direct_benchmark/ into a single Poetry
project to eliminate cross-project path dependency issues.
Changes:
- Create classic/pyproject.toml with merged dependencies from all three projects
- Remove individual pyproject.toml and poetry.lock files from subdirectories
- Update all CLAUDE.md files to reflect commands run from classic/ root
- Update all README.md files with new installation and usage instructions
All packages are now included via the packages directive:
- forge/forge (core agent framework)
- original_autogpt/autogpt (AutoGPT agent)
- direct_benchmark/direct_benchmark (benchmark harness)
CLI entry points preserved: autogpt, serve, direct-benchmark
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When a challenge times out but the agent's solution would have passed
evaluation, this is now clearly indicated:
- Completion blocks show "TIMEOUT (would have passed)" in yellow
- Recent completions panel shows hourglass icon + "would pass" suffix
- Summary table has new "Would Pass" column
- Final summary shows "+N would pass" count
- Success rate includes "would pass" challenges
The evaluator still runs on timed-out challenges to calculate the score,
but success remains False. This gives visibility into near-misses that
just needed more time.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously, the evaluator would run on all results including timed-out
challenges. If the agent happened to write a working solution before
timing out, evaluation would pass and override success=True, resulting
in contradictory output showing both PASS and "timed out".
Now we skip evaluation for timed-out challenges - they cannot pass.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The TicTacToe and other challenges use pytest-based test files for
evaluation. Without pytest installed in the benchmark virtualenv,
these evaluations were silently failing.
Root cause: test.py imports pytest but the package wasn't a dependency,
causing ModuleNotFoundError during evaluation subprocess.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Include config:challenge:attempt and timestamp in completion block
header for easier debugging and log correlation.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove selenium.py and test_selenium.py
- Add playwright_browser.py with WebPlaywrightComponent
- Update web component exports to use Playwright
- Update dependencies in pyproject.toml/poetry.lock
- Minor agent and reflexion strategy improvements
- Update CLAUDE.md documentation
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously cost was hardcoded to 0.0. Now extracts cumulative cost
from MultiProvider.get_incurred_cost() after each step execution.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Benchmarks now automatically save progress and resume from where they
left off. State is persisted to .benchmark_state.json in reports dir.
Features:
- Auto-resume: runs skip already-completed challenges
- --fresh: clear all state and start over
- --retry-failures: re-run only failed challenges
- --reset-strategy/model/challenge: selective resets
- `state show/clear/reset` subcommands for state management
- Config mismatch detection with auto-reset
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add --ci flag that disables Rich Live display while preserving
completion blocks. Auto-detects CI environment via CI env var or
non-TTY stdout. Prints progress every 10 completions for visibility.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Calculate max columns based on terminal width (up to 10)
- Reduced panel width from 35 to 30 chars to fit more
- Wider terminals can now show more parallel runs side-by-side
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fixes:
- Use run_key (config:challenge) instead of just config_name for tracking
active runs - allows multiple challenges from same config to run in parallel
- Add asyncio.sleep(0) yields to let multiple tasks acquire semaphore
and start before any proceed with work
- Always print completion blocks (not just failures) for visibility
This should properly show 8/8 active runs when running with --parallel 8.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
UI improvements:
- Multi-column layout: each active config gets its own panel showing
challenge name and step history (last 6 steps with status)
- Copy-paste completion blocks: when a challenge finishes (especially
failures), prints a detailed block with all steps for easy debugging
- Configurable logging: suppresses noisy LLM provider warnings unless
--debug flag is set
- Pass debug flag through harness to UI
Example active runs panel:
┌─ one_shot/claude ─┬─ rewoo/claude ────┐
│ ReadFile │ WriteFile │
│ ✓ #1 read_file │ ✓ #1 think │
│ ✓ #2 write_file │ ✓ #2 plan │
│ ● step 3: ... │ ● step 3: ... │
└───────────────────┴───────────────────┘
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add step callback to AgentRunner for real-time step logging
- BenchmarkUI now shows:
- Active runs with current step info
- Recent steps panel with colored config prefixes
- Proper Live display refresh (implements __rich_console__)
- Each config gets a distinct color for easy identification
- Verbose mode prints step logs immediately with config prefix
- Fix Live display not updating (pass UI object, not rendered content)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove vcrpy and pytest-recording dependencies
- Remove tests/vcr/ directory and vcr_cassettes submodule
- Remove .gitmodules (only had cassette submodule)
- Simplify CI workflow - no more cassette checkout/push/PAT_REVIEW
- Tests requiring API keys now skip if not set (fork PRs)
- Update CLAUDE.md files to remove cassette references
- Fix broken agbenchmark path in pyproject.toml
Security improvement: removes need for PAT with cross-repo write access.
Fork PRs will have API-dependent tests skipped (GitHub protects secrets).
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove old benchmark/ folder with agbenchmark framework
- Move challenges to direct_benchmark/challenges/
- Move analysis tools (analyze_reports.py, analyze_failures.py) to direct_benchmark/
- Move challenges_already_beaten.json to direct_benchmark/
- Update CI workflow to use direct_benchmark
- Update CLAUDE.md files with new benchmarking instructions
- Add benchmarking section to original_autogpt/CLAUDE.md
The direct_benchmark harness directly instantiates agents without HTTP
server overhead, enabling parallel execution with asyncio semaphore.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add NONINTERACTIVE_MODE env var support to AppConfig for disabling
user interaction during automated runs
- Benchmark harness now sets NONINTERACTIVE_MODE=True when starting agents
- Add agent configuration logging at server startup (model, strategy, etc.)
- Harness logs env vars being passed to agent for verification
- Add --agent-output flag to show full agent server output for debugging
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add model comparison support to test harness (claude, openai, gpt5, opus presets)
- Add --models, --smart-llm, --fast-llm, --list-models CLI args
- Add real-time logging with timestamps and progress indicators
- Fix success parsing bug: read results[0].success instead of non-existent metrics.success
- Fix agbenchmark TestResult validation: use exception typename when value is empty
- Fix WebArena challenge validation: use strings instead of integers in instantiation_dict
- Fix Agent type annotations: create AnyActionProposal union for all prompt strategies
- Add pytest integration tests for the strategy benchmark harness
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
<!-- Clearly explain the need for these changes: -->
This PR improves the Langfuse tracing implementation in the chat feature
by adopting the v3 SDK patterns, resulting in cleaner code and better
observability.
### Changes 🏗️
- **Simplified Langfuse client usage**: Replace manual client
initialization with `langfuse.get_client()` global singleton
- **Use v3 context managers**: Switch to
`start_as_current_observation()` and `propagate_attributes()` for
automatic trace propagation
- **Auto-instrument OpenAI calls**: Use `langfuse.openai` wrapper for
automatic LLM call tracing instead of manual generation tracking
- **Add `@observe` decorators**: All chat tools now have
`@observe(as_type="tool")` decorators for automatic tool execution
tracing:
- `add_understanding`
- `view_agent_output` (renamed from `agent_output`)
- `create_agent`
- `edit_agent`
- `find_agent`
- `find_block`
- `find_library_agent`
- `get_doc_page`
- `run_agent`
- `run_block`
- `search_docs`
- **Remove manual trace lifecycle**: Eliminated the verbose `finally`
block that manually ended traces/generations
- **Rename tool**: `agent_output` → `view_agent_output` for clarity
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Verified chat feature works with Langfuse tracing enabled
- [x] Confirmed traces appear correctly in Langfuse dashboard with tool
spans
- [x] Tested tool execution flows show up as nested observations
#### For configuration changes:
- [x] `.env.default` is updated or already compatible with my changes
- [x] `docker-compose.yml` is updated or already compatible with my
changes
- [x] I have included a list of my configuration changes in the PR
description (under **Changes**)
No configuration changes required - uses existing Langfuse environment
variables.
When tool calls fail validation, the error messages now include:
- What arguments were actually provided
- The expected parameter schema with types and required/optional indicators
This helps LLMs understand and fix their mistakes when retrying,
rather than just being told a parameter is missing.
Example improved error:
Invalid function call for write_file: 'contents' is a required property
You provided: {"filename": 'story.txt'}
Expected parameters: {"filename": string (required), "contents": string (required)}
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add WebFetchComponent for fast HTTP-based page fetching without browser
overhead. Uses trafilatura for intelligent content extraction.
Commands:
- fetch_webpage: Extract main content as text/markdown/xml
- Removes navigation, ads, boilerplate automatically
- Extracts page metadata (title, description, author, date)
- Extracts and lists page links
- Much faster than Selenium-based read_webpage
- fetch_raw_html: Get raw HTML for structure inspection
- Optional truncation for large pages
Features:
- Trafilatura-powered content extraction (best-in-class accuracy)
- Automatic link extraction with relative URL resolution
- Page metadata extraction (OG tags, meta tags)
- Configurable timeout, max content length, max links
- Proper error handling for timeouts and HTTP errors
- 19 comprehensive tests
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add generate_block_docs.py script that introspects block code to
generate markdown
- Support manual content preservation via <!-- MANUAL: --> markers
- Add migrate_block_docs.py to preserve existing manual content from git
HEAD
- Add CI workflow (docs-block-sync.yml) to fail if docs drift from code
- Add Claude PR review workflow (docs-claude-review.yml) for doc changes
- Add manual LLM enhancement workflow (docs-enhance.yml)
- Add GitBook configuration (.gitbook.yaml, SUMMARY.md)
- Fix non-deterministic category ordering (categories is a set)
- Add comprehensive test suite (32 tests)
- Generate docs for 444 blocks with 66 preserved manual sections
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
<!-- Clearly explain the need for these changes: -->
### Changes 🏗️
<!-- Concisely describe all of the changes made in this pull request:
-->
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Extensively test code generation for the docs pages
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Introduces an automated documentation pipeline for blocks and
integrates it into CI.
>
> - Adds `scripts/generate_block_docs.py` (+ tests) to introspect blocks
and generate `docs/integrations/**`, preserving `<!-- MANUAL: -->`
sections
> - New CI workflows: **docs-block-sync** (fails if docs drift),
**docs-claude-review** (AI review for block/docs PRs), and
**docs-enhance** (optional LLM improvements)
> - Updates existing Claude workflows to use `CLAUDE_CODE_OAUTH_TOKEN`
instead of `ANTHROPIC_API_KEY`
> - Improves numerous block descriptions/typos and links across backend
blocks to standardize docs output
> - Commits initial generated docs including
`docs/integrations/README.md` and many provider/category pages
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
631e53e0f6. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
- Extract _get_tool_error_message helper method
- Replace 20+ levels of nesting with simple for loop
- Improve readability of tool_result construction
- Update benchmark poetry.lock
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Replace basic DuckDuckGo-only search with a modern tiered system:
1. Tavily (primary) - AI-optimized results with content extraction
- AI-generated answer summaries
- Relevance scoring
- Full page content extraction via search_and_extract command
2. Serper (secondary) - Fast, cheap Google SERP results
- $0.30-1.00 per 1K queries
- Real Google results without scraping
3. DDGS multi-engine (fallback) - Free, no API key required
- Automatic fallback chain: DuckDuckGo → Bing → Brave → Google → etc.
- 8 search backends supported
Key changes:
- Upgrade duckduckgo-search to ddgs v9.10 (renamed successor package)
- Add Tavily and Serper API integrations
- Implement automatic provider selection and fallback chain
- Add search_and_extract command for research with content extraction
- Add TAVILY_API_KEY and SERPER_API_KEY to env templates
- Update benchmark httpx constraint for ddgs compatibility
- 23 comprehensive tests for all providers and fallback scenarios
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Change Agent class to use ActionProposal instead of OneShotAgentActionProposal
to support multiple prompt strategy types
- Widen display_thoughts parameter type from AssistantThoughts to ModelWithSummary
- Fix speak attribute access in agent_protocol_server with hasattr check
- Add type: ignore comments for intentional thoughts field overrides in strategies
- Remove unused OneShotAgentActionProposal import
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix 32+ flake8 E501 (line too long) errors by shortening descriptions
- Remove unused import in todo.py
- Fix test_todo.py argument order (config= keyword)
- Add type annotations to fix pyright errors where straightforward
- Add noqa comments for flake8 false positives in __init__.py
- Remove unused nonlocal declarations in main.py
- Run black and isort to fix formatting
- Update CLAUDE.md with improved linting commands
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously, when a user selected "Once" or "Always" with feedback (via Tab),
the command was NOT executed because UserFeedbackProvided was raised before
checking the approval scope. This fix changes the architecture from
exception-based to return-value-based.
Changes:
- Add PermissionCheckResult class with allowed, scope, and feedback fields
- Change check_command() to return PermissionCheckResult instead of bool
- Update prompt_fn signature to return (ApprovalScope, feedback) tuple
- Add pending_user_feedback mechanism to EpisodicActionHistory
- Update execute() to handle feedback after successful command execution
- Feedback message explicitly states "Command executed successfully"
- Add on_auto_approve callback for displaying auto-approved commands
- Add comprehensive tests for approval/denial with feedback scenarios
Behavior:
- Once + feedback → Execute command, then send feedback to agent
- Always + feedback → Execute command, save permission, send feedback
- Deny + feedback → Don't execute, send feedback to agent
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Adds a custom Rich-based interactive selector for the command approval
workflow. Features include:
- Arrow key navigation for selecting approval options
- Tab to add context to any selection (e.g., "Once + also check file x")
- Dedicated inline feedback option with shadow placeholder text
- Quick select with number keys 1-5
- Works within existing asyncio event loop (no prompt_toolkit dependency)
Also adds UIProvider abstraction pattern for future UI implementations.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
