Build a parent/child tree: crew_start ├── task_start │ ├── agent_start │ └── agent_complete (started=agent_start)

Add a tool result to the cache. Args: tool: Name of the tool. input: Input string used for the tool. outp

Handle checkpoint config for a kickoff method. If *from_checkpoint* carries a ``restore_from`` path, builds and returns a restored instance (

Return ``os.environ`` augmented with UV_INDEX_* credentials for every private index referenced under ``[tool.uv.sources]`` in ``pyproject.toml``.

Serialize a module-level callable as a ``"module.qualname"`` string. Args: fn: The callable to serialize. Returns: The dotte

Compute a weighted composite relevance score from semantic, recency, and importance. composite = w_semantic * semantic + w_recency * decay + w_im

Create an agent executor for the agent. Returns: An instance of the CrewAgentExecutor class.

Creates a ChromaDB client from configuration.

Create a standardized panel with consistent styling.

Create a minimal valid PDF for testing.

Return the path for SQLite database / app-data storage. Creates the directory if it does not exist.

Deploy a crew using either UUID or project name. Args: uuid (Optional[str]): The UUID of the crew to deploy.

Execute an A2A task using a CrewAI agent. Args: agent: The CrewAI agent to execute the task. context: The A2A request context con

Emit a final ``FlowFinishedEvent`` and finalize the trace batch. Pairs with ``flow.defer_trace_finalization = True`` (or ``Conversati

Fork a Crew from a checkpoint, creating a new execution branch. Args: config: Checkpoint configuration with ``restore_from`` set.

Format skill information for agent prompt injection. At METADATA level: returns name and description only. At INSTRUCTIONS level or above: re

Get or create a page for the specified thread.

Get or create a page for the specified thread.

Return the access token if valid and not expired, else None.

Get Gemini file uploader.

Handle context length exceeded by either summarizing or raising an error. Args: respect_context_window: Whether to respect context window

Pydantic-compatible function to import a class/function from a string path. Args: v: Import path string in the format 'module.path.ClassN

Check if batch is initialized

Return ``(is_newer, current_version, latest_version)``.

List records in a scope, newest first. Args: scope: Optional scope path prefix to filter by. If None and root_scope

Load state and pending feedback context. Args: flow_uuid: Unique identifier for the flow instance Returns: T

Load one skill input into Skill objects. Accepts a pre-loaded Skill object, skill search path, inline SKILL.md string, or '@org/name' registr

Load the most recent state for a given flow UUID. Args: flow_uuid: Unique identifier for the flow instance Returns:

Mark a todo as running by step number.

Memoize a method by caching its results based on arguments. Handles both sync and async methods. Pydantic BaseModel instances are converted t

Parse a ``--inputs`` JSON object, exiting with a pointed error if invalid.

Read a TOML file from disk and return its parsed contents.

Register an event handler for a specific event type. Args: event_type: The event class to listen for handler: The han

Save the access token and its expiration time. Args: access_token: The access token to save. expires_at: The UNIX tim

Set the ID of the triggering event for causal chain tracking. Args: event_id: The event_id that triggered the current execution, or None.

Gracefully shutdown the event loop and wait for all tasks to finish. Args: wait: If True, wait for all pending tasks to complete

Store files for a crew execution. Args: execution_id: Unique identifier for the crew execution. files: Dictionary mapping names t

Iterate over frames, optionally filtered by channel.

Unregister a specific global after_tool_call hook. Args: hook: The hook function to remove Returns: True if the hook was fou

Unregister a specific global before_tool_call hook. Args: hook: The hook function to remove Returns: True if the hook was fo

Update the [project] version field in a pyproject.toml file. Args: file_path: Path to pyproject.toml file. new_version: New versi

Validate JSON crew structure without kicking off the crew.

Import and return the class/function from the import path. Args: v: Import path string in the format 'module.path.ClassName'. Return

Factory function that returns crewai platform tools. Args: apps: List of platform apps to get tools that are available on the platform.

Add A2A delegation instructions to prompt. Args: a2a_agents: Dictionary of A2A agent configurations task_description: Original ta

Build an UploadResult for a completed upload. Args: file_id: The uploaded file ID. content_type: The file's content t

Clear the read and write streams.

Helper to create a skill directory with SKILL.md.

Extract raw finish_reason and response_id from an Anthropic ``Message`` / ``BetaMessage``. Anthropic exposes ``stop_reason`` (e.g. ``"

Extract token usage and response metadata from OpenAI ChatCompletion.

Extract token usage and response metadata from Responses API response.

Extract token usage and response metadata from Gemini response.

Fetch MCP server configurations via CrewAI+ API. Sends a GET request to the CrewAI+ mcps/configs endpoint with comma-separated slugs.

Format messages for Azure AI Inference API. Args: messages: Input messages Returns: List of dict objects wit

Get or create the OpenAI client.

Get the appropriate client - instance-specific or global.

Get the appropriate purpose for a file based on content type. OpenAI Files API requires different purposes for different file types: - Images

Platform-appropriate per-user credential directory (mode 0o700).

Handles the Crew planning.

Initialize backend batch and send collected events.

Invoke before_llm_call hooks for direct LLM calls (no agent context). This method should be called by native provider implementations before

Type guard to check if the client is an asynchronous AsyncQdrantClient. Args: client: The client to check. Returns: True if

Type guard to check if the client is an asynchronous AsyncClientAPI. Args: client: The client to check. Returns: True if the

Type guard to check if the client is a synchronous QdrantClient. Args: client: The client to check. Returns: True if the cli

Type guard to check if the client is a synchronous ClientAPI. Args: client: The client to check. Returns: True if the client

Create a mock MCPClient that returns *tool_definitions*.

Build an in-memory .tar.gz containing the given filename → content mapping.

True when a crew runs inside a flow session (context or batch ownership).

Parse an AMP reference into *(slug, optional tool name)*. Accepts both bare slugs (``"notion"``, ``"notion#search"``) and the legacy

Parse the LLM response into a StepObservation. The LLM may return: - A StepObservation instance directly (streaming + litellm path)

Read raw bytes from a secure file, or None if absent.

Subclass must implement: transform validated params dict into API request params.

Return the bound ``Memory`` or raise a clear error if missing.

Coerce a checkpoint field value. Returns: CheckpointConfig — use this config. _SENTINEL — explicit opt-out (``False``), stop walk

Emit agent execution event. Args: formatted_answer: Agent's response to log.