Collapse free-form feedback to a predefined outcome using LLM. This method uses the specified LLM to interpret the human's feedback a

Extract token usage and response metadata from Anthropic response.

Format messages with role labels for summarization. Skips system messages. Handles None content, tool_calls, and multimodal content blocks.

Get or create a code interpreter for the specified thread. Args: thread_id: Thread ID for the code interpreter session R

Return ``True`` if *fn* cannot survive a serialize/deserialize round-trip. Built-in functions, plain module-level functions, and classes produce

Read the user-data JSON file, returning ``{}`` on missing/corrupt.

Prepare documents for ChromaDB by extracting IDs, texts, and metadata. Args: documents: List of BaseRecord documents to prepare. Ret

Run an async coroutine from sync context without blocking event loop.

Run a void telemetry procedure, swallowing failures.

Set status for an explicit plan step reported by the planner.

Asynchronously emit an event to registered async handlers. Only processes async handlers. Use in async contexts. Args: s

Clean up Stagehand resources.

Kill the cached persistent sandbox if one exists.

Return the configured declarative flow source for flow projects.

Create a new collection in Qdrant. Keyword Args: collection_name: Name of the collection to create. Must be unique. v

Extract output from code interpreter response stream. Args: response: Response from code interpreter execution Returns: Extr

Extract tool call ID, name, and arguments from various provider formats. Args: tool_call: The tool call object to extract info from.

Restore a Crew from a checkpoint, ready to resume via kickoff(). Args: config: Checkpoint configuration with ``restore_from`` set

Get all registered global after_tool_call hooks. Returns: List of registered after hooks

Get the current parent event ID from the stack.

Initialize a new trace batch (thread-safe)

Return ``(yanked, reason)`` for the currently installed version.

Check if tracing is enabled in current execution context. Returns: True if tracing is enabled in context, False otherwise. Return

Return ``message`` with the cache-breakpoint flag set. Returns a new dict so callers can safely pass literal dicts.

Mark a todo as completed by step number.

Push an event ID and type onto the scope stack.

Register a global hook for an interception point.

Save the data to the specified file using pickle. Args: data: The data to be saved to the file.

Convert a dotted path string to the callable it references. If *value* is already callable it is returned as-is, with a warning if it cannot

Generate a standard prompt for task execution. Returns: A dictionary containing the constructed prompt(s).

Validate an audio file against constraints. Args: file: The audio file to validate. constraints: Audio constraints to validate ag

Common payload refinement for all tools.

Helper to create a skill directory with SKILL.md.

Emit one fake ``LLMCallCompletedEvent`` with ``current_flow_id`` pinned to ``flow_id``. Runs in a freshly-copied context so the value the bus

Fetch action schemas from the platform API.

Format messages for OpenAI API.

Format messages for Converse API following AWS documentation. Note: Returns dict[str, Any] instead of LLMMessage because Bedrock uses

Get or create the Gemini client.

Get OpenAI client parameters.

Stable, privacy-preserving machine fingerprint (cross-platform).

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

Invoke step callback if configured. Args: formatted_answer: Current agent response.

Return True if the cache is less than 24 hours old.

Make a JSON-RPC 2.0 MCP request to Agent Handler.

Whether a 400 is OpenAI refusing `reasoning_effort` alongside tools. GPT-5.6 applies a server-side `reasoning_effort` default and then reject

Import a module and walk attribute lookups to resolve a dotted path. Handles multi-level qualified names like ``"module.ClassName.method"`` b

Split messages into chunks at message boundaries. Excludes system messages from chunks. Each chunk stays under max_tokens based on estimated

Translate a chat-format message into Responses ``input`` items. Tool calling is expressed differently by the two APIs. Chat Completions

Track token usage internally in the LLM instance. Args: usage_data: Token usage data from the API response

Validate the search query to ensure it's safe to execute. Only SELECT and SHOW statements are allowed for security reasons. Args:

Cancel streaming and clean up resources.

Promote a skill to INSTRUCTIONS disclosure level. Idempotent: returns the skill unchanged if already at or above INSTRUCTIONS. Args:

Add content to the knowledge base.

Split a text file into chunks of maximum character count. Yields chunks one at a time to minimize memory usage. Args: file: The text

Creates a deep copy of the Crew instance. Returns: Crew: A new instance with copied components

Delete a collection and all its data. Keyword Args: collection_name: Name of the collection to delete. Raises:

Detect MIME type from file content. Uses python-magic if available for accurate content-based detection, falls back to mimetypes module using

Force additionalProperties=false on all object-type dicts recursively. OpenAI strict mode requires all objects to have additionalProperties=false

Build an expression with the standard Flow runtime context.

Generate a UUID-based filename with extension from content type. Args: content_type: MIME type to derive extension from. Returns:

Get merged crew and task files. Task files override crew files with the same name. Args: crew_id: Unique identifier for the crew exe

Get the current crew context from OpenTelemetry baggage. Returns: CrewContext instance containing crew context information, or None if no

Return the generated-project dependency for CrewAI with tools.

Get an existing collection or create it if it doesn't exist. Keyword Args: collection_name: Name of the collection to get or crea

Extract tool call streaming events from mock emit calls.

Get the ID of the event that triggered the current execution. Returns: The event_id of the triggering event, or None if not in a triggere

Check if this cached upload has expired.

Normalize a scope path by removing double slashes and ensuring proper format. Args: path: The raw scope path (e.g. '/crew/MyCrewName//age

Parse '@org/skill-name' into (org, name), ignoring any version pin. Retained for backwards compatibility. Use :func:`parse_skill_ref` to read

Recall relevant memories. Delegates to this flow's memory. Args: query: Natural language query. **kwargs: Passed to m

Request human input during LLM hook execution. This method pauses live console updates, displays a prompt to the user, waits for thei

Recursively resolve all local $refs in the given JSON Schema using $defs as the source. This is needed because Pydantic generates $ref-based sche

Set the platform integration token in the current context. Args: integration_token: The integration token to set.

Strip JSONC comments and trailing commas while preserving string values.

Remove format annotations that OpenAI strict mode doesn't support. OpenAI only supports: date-time, date, time, duration. Other formats like

Check if the model supports function calling. Note: This method is only used by the litellm fallback path. Native providers override

Check if the model supports stop words. Note: This method is only used by the litellm fallback path. Native providers override this m

Context manager to suppress all warnings. Yields: None during the context execution. Note: This implementation consolidates

Get all available agent tools

Support instance methods by implementing the descriptor protocol. Args: obj: The instance that the method is accessed through.