Append new training data for a specific agent and iteration. Args: train_iteration: The training iteration number. ag

Retrieve an auth scheme by key.

Execute the crew's workflow. Args: inputs: Optional input dictionary for task interpolation. input_files: Optional di

Sync wrapper for aget.

Emit an event to all registered handlers. If handlers have dependencies (registered with depends_on), they execute in dependency orde

High-level LLM call method. Args: messages: Input messages for the LLM. Can be a string or list of message d

Build a runnable declarative flow from contents or a file path.

Print to console. Simplified to only handle panel-based output.

Store an auth scheme.

Main method for tool execution.

Update an existing memory record by ID. Args: record_id: ID of the record to update. content: New content; re-embedde

Print ``content`` with optional color, honoring suppression context.

Executes the tool synchronously. Args: *args: Positional arguments for the tool. **kwargs: Keyword arguments for the

Context manager for temporary event handling scope. Useful for testing or temporary event handling. All handlers registered within th

Read the file content as string.

Return keys for dict unpacking.

Return scope info for path. Args: path: Scope path to get info for. If None and root_scope is set, defaults to ro

Execute the agent with the given messages using the AgentExecutor. This method provides standalone agent execution without requiring a Crew.

Block until all pending event handlers complete. This method waits for all futures from previously emitted events to finish executing

Read the file content as bytes.

Generic handler for action events (LLM calls, tool usage). Args: event_type: Type of the event. source: Source object

Return a NL2SQLTool wired to an in-memory SQLite DB. Schema-introspection is patched out so we can create the tool without a real PostgreSQL

Async high-level LLM call method. Args: messages: Input messages for the LLM. Can be a string or list of mes

Print a panel with consistent formatting if verbose is enabled.

Create a Pydantic model from a JSON schema. This function takes a JSON schema as input and dynamically creates a Pydantic model class based o

Native async kickoff method using async task execution throughout. Unlike kickoff_async which wraps sync kickoff in a thread, this method

Sanitize tool name for LLM provider compatibility. Normalizes Unicode, splits camelCase, lowercases, replaces invalid characters with undersc

Add an event to the record and wire its edges. Args: event: The event to insert. Returns: The created node.

Raise ValueError if *sql_query* is not permitted under the current config. Splits the query on semicolons and validates each statement

Format this match as a human-readable string including metadata. Returns: A multi-line string with score, content, categories, an

Create an AgentExecutor without validation — for unit tests. Uses model_construct to skip Pydantic validators so plain Mock() objects are acc

Log a message with timestamp if verbose mode is enabled. Args: level: The log level (e.g., 'info', 'warning', 'error').

Connect to MCP server and initialize session. Returns: Self for method chaining. Raises: ConnectionError: If

Retrieve a prompt slice by key. Args: slice: The key of the prompt slice to retrieve. Returns: The prompt sl

Remove all nodes from the record under the write lock.

Return the static Flow Definition built from this Flow class.

Convert MCP server references/configs to CrewAI tools.

Call the LLM with the given messages. Args: messages: Input messages for the LLM. Can be a string or list of

Run all checks. Later checks are skipped when earlier ones make them impossible (e.g. no pyproject.toml → no lockfile check).

Load the data from the specified file using pickle. Returns: The data loaded from the file.

Execute the update mechanism and return result. Args: client: A2A client instance. message: Message to send.

Emit LLM call completed event.

Count records in scope (and subscopes). Args: scope_prefix: Optional scope path (None = all). Returns: Numbe

Execute a task with the agent. Args: task: Task to execute. context: Context to execute the task in. tool

Asynchronous kickoff method to start the crew execution. Args: inputs: Optional input dictionary for task interpolation.

Parse {% js 'script.js' %} tag. Args: parser: Jinja2 parser instance. Returns: Output node with rendered scr

Append a user message, run one conversational turn, and return output. .. warning:: **EXPERIMENTAL.** This is the public entry po

Load a ``Crew`` from a JSON/JSONC definition file. The definition file describes the crew's agents, tasks, process type, and default inputs.

Execute the task synchronously.

Register an extension. Args: extension: The extension to register.

Save current settings to settings.json.

Store a single item in memory (synchronous). Routes through the same serialized save pool as ``remember_many`` to prevent races, but

Validate that a URL is safe to fetch. Blocks ``file://`` scheme entirely. For ``http``/``https``, resolves DNS and checks that the target IP

Interpolate placeholders (e.g., {key}) in a string while leaving JSON untouched. Only interpolates placeholders that follow the pattern {variable_

Reset the last event ID to None. Should be called at the start of a new flow or when resetting event state.

Return the final result after stream exhaustion or completion.

Execute telemetry operation safely, checking both readiness and environment variables. Args: operation: A callable that performs

Retrieve a cached tool result. Args: tool: Name of the tool. input: Input string used for the tool. Returns:

Build a visualization structure projection from a FlowDefinition.

Creates or returns an LLM instance based on the given llm_value. Args: llm_value: LLM instance, model name string, config dict, None, or

Generate JSON schema description of a Pydantic model. This function takes a Pydantic model class and returns its JSON schema, which includes

Create a simple analyst crew for file analysis.

Track token usage from Bedrock response.

Request input from the user during flow execution. Blocks the current thread until the user provides input or the timeout expires. Wo

Convert CrewAI tools to OpenAI function calling format. This function converts CrewAI BaseTool and CrewStructuredTool objects into the OpenAI

Generates the task prompt with optional markdown formatting. When the markdown attribute is True, instructions for formatting the res

Apply stop words to truncate response content. This method provides consistent stop word behavior across all native SDK providers. Na

Build a multimodal message with text and file content.

Build a ``requests.Response``-like mock with the attributes used by ``_make_request``.

Sync wrapper for aremove.

Reset the emission sequence counter to 1. Resets for the current context only.

Return the call-scoped streaming mode for this instance.

Creates a deep copy of the Task while preserving its original class type. Args: agents: List of agents available for the task.

Get all registered global before_llm_call hooks. Returns: List of registered before hooks