Run ruff format on agent_tui.py, cli.py, executor.py. Fix agents_dir
argument type: pass Path object instead of str to match the
load_agent_from_definition signature (Path | None).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove redundant local `import asyncio` in executor.py that caused
ruff F823 (local variable referenced before assignment)
- Clear progress state before creating Live display (fixes flash)
- Use threshold-based passed in _save_run_results so persisted results
match CLI output
- Pass agents_dir to load_agent_from_definition in _train_new_agents
so coworker references resolve correctly
- Deduplicate verbose/non-verbose benchmark execution blocks into
single context-manager expression
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1. Summary tags: Reverse the logic — for models like gpt-4.1 that wrap
their actual response in <summary> tags (with thinking/CoT before it),
extract the inner content instead of stripping it. Streaming uses a
preflight buffer that waits for <summary>; if none appears, flushes
everything normally.
2. TUI autocomplete: Change @mention accept key from Tab to right-arrow
so autocomplete doesn't steal focus from the input widget. Only
triggers when there's an active mention context with matches.
3. Tool output: Truncate tool results >4000 chars in LLM message history
to prevent the model from echoing full file contents. Add soul-layer
instruction telling the agent to summarize tool results rather than
repeating them verbatim.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add null check after _load_agent() in benchmark runner (agent can return None on circular refs)
- Escape user-sourced content in Rich markup via _safe_render() in memory panel and skills list
- Default to passed=True when benchmark case has neither expected nor criteria
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove dead `env_vars.get("MODEL")` check in _setup_env (always truthy
since MODEL is set two lines above)
- Fix test_sync_delegation mock: use return_value instead of side_effect
list and disable planning to prevent StopAsyncIteration on Python 3.10
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Introduced a case_timeout parameter in the benchmark and test functions to allow dynamic timeout settings.
- Updated the project configuration template to include a default case_timeout value of 90 seconds.
- Enhanced the handling of timeouts in benchmark results to reflect the configured case_timeout.
- Introduced a new judge tool for submitting evaluation scores with structured parameters.
- Added a function to parse judge results from various response formats.
- Updated the benchmark command to handle iterations more effectively, allowing configuration from the command line or config file.
- Implemented a method to save run results to a JSON file for better tracking of test outcomes.
- Enhanced progress display to show current iteration during benchmark runs.
- Updated project configuration template to clarify test iteration settings.
- Updated ChatTextArea to support @mention autocomplete using Tab for completion.
- Added MentionChanged message to handle autocomplete state changes.
- Improved user experience by displaying a hint for available mentions.
- Enhanced error handling in AgentTUI for agent message timeouts.
- Updated rendering logic to ensure proper display of system messages with Rich markup.
- Move progress.start() inside the try block so the finally clause
never calls progress.stop() on an un-started display
- Replace asyncio.run() with new_event_loop/run_until_complete/loop.close()
pattern, consistent with _test_new_agents and _train_new_agents
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- write_env_file: remove .upper() to preserve original key case
- load_env_vars: strip surrounding single/double quotes from values
- constants.py: fix Ollama key_name from OPENAI_API_BASE to OLLAMA_HOST
- _test_new_agents: replace asyncio.run() loop with new_event_loop + run_until_complete
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- benchmark verbose path: pass on_progress callback the same way as
the non-verbose path (was missing entirely)
- _train_new_agents: replace per-case asyncio.run() with a single
event loop (new_event_loop / run_until_complete / close) to avoid
creating and destroying a loop on every case iteration
- format_results_table: use case_index + 1 so the '#' column is
1-based, matching the display in _test_new_agents failed output
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Replace S101 assert guards with explicit if/raise RuntimeError in
benchmark.py and cli.py (3 locations)
- Fix test_create_llm_from_env_with_unaccepted_attributes to use
DEFAULT_LLM_MODEL with clear=True so the assertion isn't brittle
against the hardcoded model name
- Add n_iterations loop to _test_new_agents (was unused, now mirrors
_train_new_agents iteration pattern)
- Consolidate dotenv loading in cli.py and agent_tui.py to use the
existing load_env_vars() from utils.py instead of duplicating logic
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add missing type annotations to benchmark.py context managers, event
handlers, LoadedCases iteration methods, and fix union-attr on BaseLLM.
Fix no-any-return errors in agent_tui.py and make action_quit async to
match the Textual App supertype. Add type annotations to
_BenchmarkLiveProgress methods in cli.py and fix icon redefinition.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Format: auto-reformat agent_tui.py, benchmark.py, coworker_tools.py via ruff
- Lint: 0 remaining errors after format pass
- Mypy: fix _NullPrinter to subclass Printer for type compatibility in
executor.py, planning.py, and skill_builder.py; add isinstance(r, Message)
guards in spawn_tools.py; annotate return types and fix dict type params
and MCPToolResolver logger type in new_agent.py; add missing printer args
to get_llm_response calls
- cli.py: fix _read_config to use sentinel so falsy values (0, false) are
returned correctly instead of being treated as missing keys
- create_agent.py: replace regex-based JSONC comment stripper with a
token-aware parser that preserves // inside quoted strings (e.g. URLs)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Extract `_strip_jsonc` as the single shared helper in `create_agent.py`,
replacing the three duplicate implementations in `agent_tui.py`,
`benchmark.py`, and the inline regex in `cli.py::_read_config`.
- Apply `_strip_jsonc` (including trailing-comma removal) inside
`_read_config` so JSONC config.json files are parsed correctly.
- Add `if progress is not None:` guard inside `_make_progress_cb._cb`
to prevent a `NoneType` call when running in verbose mode.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Introduced a new `ChatTextArea` class to enhance multiline chat input functionality, allowing users to submit messages with Enter and insert newlines with Shift+Enter.
- Updated the TUI layout to replace the previous input method with `ChatTextArea`, improving user experience during chat interactions.
- Removed unused sidebar actions and adjusted input row styling for better visual consistency.
These changes aim to streamline chat interactions within the CrewAI framework, providing a more intuitive input experience.
- Introduced a `verbose` flag in the CLI for the `test` and `benchmark` commands to enable detailed logging of agent execution, including tool calls and LLM responses.
- Updated the `_run_model_benchmark` and `_test_new_agents` functions to accept the `verbose` parameter, allowing for enhanced debugging during benchmark runs.
- Implemented a `verbose_benchmark_output` context manager to manage logging output when verbose mode is enabled, improving the visibility of agent interactions.
These changes enhance the debugging capabilities of the CrewAI framework, providing users with more insights during testing and benchmarking processes.
- cli.py: use s.get('done',0)+1 instead of max(s['done'], event['case_index']) for correct progress counting
- cli.py: use explicit 'is not None' check for config_threshold to avoid treating 0.0 as falsy
- cli.py: remove unused agent_count variable
- constants.py + create_agent.py: add key_name to ollama ENV_VARS entry so API_BASE is correctly saved to OPENAI_API_BASE
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Rename suppress_benchmark_output → SuppressBenchmarkOutput and
artifacts_sandbox → ArtifactsSandbox (N801 CapWords), and drop unused
loop variable to use dict.values() (PERF102).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- B904: raise KeyboardInterrupt from err in cli_provider.py
- mypy: add TYPE_CHECKING import for SQLiteConversationStorage, annotate
_initialized class var in TaskScheduler, fix Match type params and
Returning Any in create_agent.py
- tests: mock aget_llm_response in 3 integration tests that fail when
network is blocked but OPENAI_API_KEY is set
- flow.py: use asyncio.run_coroutine_threadsafe() instead of asyncio.run()
when a loop is already running in ask() and say()
- cli.py: fix threshold=0.0 treated as falsy by using `is not None` check
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Added a `CreateRoomScreen` modal for creating new rooms with agent selection and engagement options.
- Updated the main TUI layout to include a sidebar for room management, allowing users to create and switch between rooms.
- Enhanced the configuration handling to support room definitions and engagement modes.
- Refactored existing code to accommodate new room functionalities and improve overall structure.
These changes enhance the user experience by enabling better organization and interaction with multiple agents in the CrewAI framework.
- Introduced a new `LoadedCases` class to encapsulate benchmark cases and optional thresholds, improving data management.
- Updated `load_benchmark_cases` function to support loading cases from both bare arrays and object wrappers with a threshold.
- Modified CLI options to allow dynamic threshold configuration, defaulting to a value from `config.json` if not specified.
- Enhanced error handling for invalid benchmark case formats and added tests to validate new functionality.
These changes aim to improve the flexibility and usability of benchmark case management within the CrewAI framework.
- Added functionality to load environment variables from a `.env` file if it exists, improving configuration management.
- Updated the CLI to fallback to a `benchmarks` directory for test cases if the `tests` directory is not found, ensuring compatibility with previous project structures.
- Refactored benchmark case path handling to streamline testing processes.
These changes aim to improve the usability and flexibility of the CrewAI CLI in various project setups.
- Added a `_safe_render` function to escape Rich markup and convert markdown to Rich format.
- Implemented token-by-token streaming for agent responses in the TUI, improving user experience during interactions.
- Updated the CLI to allow selection of LLM providers and models, enhancing flexibility in agent creation.
- Refactored benchmark case paths to use a `tests` directory instead of `benchmarks`.
- Introduced a `last_stream_result` property in the `NewAgent` class to retrieve the latest streaming response.
These changes aim to provide a more interactive and user-friendly experience in managing agents within the CrewAI framework.
- Introduced a new `create_agent` command for interactive agent definition.
- Added `agent_tui.py` for a conversational TUI supporting multi-agent interactions.
- Updated CLI to support agent creation and training workflows.
- Enhanced `.gitignore` to exclude demo files and configuration artifacts.
- Implemented a benchmark runner for testing agent performance against defined cases.
This commit lays the groundwork for a more interactive and user-friendly experience in managing agents within the CrewAI framework.
