[`agentic_crew.crew`](#module-agentic_crew.crew)
CrewAI Entry Point
This file is the standard CrewAI entry point that crewai run expects.
It loads configuration from crewbase.yaml and executes the crew.
Module Contents
Section titled “Module Contents”Classes
Section titled “Classes”CrewAgents | Wrapper class for CrewAI crew execution. |
|---|
Functions
Section titled “Functions”get_crewbase_path | Get path to crewbase.yaml. |
|---|---|
load_crewbase | Load crewbase.yaml configuration. |
kickoff | Main entry point for crewai run. |
train | Train the crew using memory/learning features. |
agentic_crew.crew.get_crewbase_path() → pathlib.Path
Section titled “agentic_crew.crew.get_crewbase_path() → pathlib.Path”Get path to crewbase.yaml.
class agentic_crew.crew.CrewAgents(config_path: pathlib.Path | None = None)
Section titled “class agentic_crew.crew.CrewAgents(config_path: pathlib.Path | None = None)”Wrapper class for CrewAI crew execution.
Provides a class-based interface for flows to instantiate and run crews. This bridges the gap between Flow expectations and the YAML-based crew config.
Initialization
Section titled “Initialization”Initialize CrewAgents with optional custom config path.
property config : dict[str, Any]
Section titled “property config : dict[str, Any]”Lazy load configuration.
property crew : crewai.Crew
Section titled “property crew : crewai.Crew”Lazy load crew from YAML.
_get_crew_for_task(task_name: str | None = None) → crewai.Crew
Section titled “_get_crew_for_task(task_name: str | None = None) → crewai.Crew”Get a crew instance, optionally filtered to a specific task.
When task filtering is needed, creates a fresh crew instance to avoid mutating the cached crew’s task list.
Args: task_name: Optional task name to filter to
Returns: Crew instance (cached if no filtering, fresh if filtering)
Raises: ValueError: If task_name is provided but not found
kickoff(inputs: dict[str, Any] | None = None) → Any
Section titled “kickoff(inputs: dict[str, Any] | None = None) → Any”Execute the crew with given inputs.
Args: inputs: Optional dict with:
- task: Specific task name to run (filters crew tasks)
- Any other inputs passed to crew.kickoff()
Returns: CrewOutput from crew execution
Note: This method does NOT mutate the inputs dict passed by caller.
kickoff_async(inputs: dict[str, Any] | None = None) → Any
Section titled “kickoff_async(inputs: dict[str, Any] | None = None) → Any”Async version of kickoff for parallel execution.
Note: This method does NOT mutate the inputs dict passed by caller.
agentic_crew.crew.load_crewbase()
Section titled “agentic_crew.crew.load_crewbase()”Load crewbase.yaml configuration.
agentic_crew.crew.kickoff(inputs: dict[str, Any] | None = None)
Section titled “agentic_crew.crew.kickoff(inputs: dict[str, Any] | None = None)”Main entry point for crewai run.
Args: inputs: Optional dict with:
- task: Specific task name to run
- All other task inputs from crewbase.yaml
Note: This function does NOT mutate the inputs dict passed by caller.
agentic_crew.crew.train(n_iterations: int = 5, inputs: dict[str, Any] | None = None)
Section titled “agentic_crew.crew.train(n_iterations: int = 5, inputs: dict[str, Any] | None = None)”Train the crew using memory/learning features.