agentic_fabric.core.loader¶
Loader module - creates CrewAI Crew objects from YAML configurations.
Note: This module imports CrewAI lazily to allow the core package to be installed without requiring CrewAI. For framework-agnostic usage, prefer using the decomposer module with runners instead.
Module Contents¶
Functions¶
Load knowledge sources from the specified paths. |
|
Create an Agent from YAML configuration. |
|
Create a Task from YAML configuration. |
|
Load a complete Crew from configuration. |
Data¶
API¶
- agentic_fabric.core.loader.logger = 'getLogger(...)'¶
- agentic_fabric.core.loader.load_knowledge_sources(knowledge_paths: list[pathlib.Path]) list¶
Load knowledge sources from the specified paths.
Args: knowledge_paths: List of paths to knowledge directories.
Returns: List of TextFileKnowledgeSource objects.
Raises: ImportError: If crewai is not installed.
- agentic_fabric.core.loader.create_agent_from_config(agent_name: str, agent_config: dict, tools: list | None = None) crewai.Agent¶
Create an Agent from YAML configuration.
Args: agent_name: Name/key of the agent. agent_config: Agent configuration dict with role, goal, backstory. tools: Optional list of tools to give the agent.
Returns: Configured Agent instance.
Raises: ImportError: If crewai is not installed.
- agentic_fabric.core.loader.create_task_from_config(task_name: str, task_config: dict, agent: crewai.Agent) crewai.Task¶
Create a Task from YAML configuration.
Args: task_name: Name/key of the task. task_config: Task configuration dict with description, expected_output. agent: Agent to assign to this task.
Returns: Configured Task instance.
Raises: ImportError: If crewai is not installed.
- agentic_fabric.core.loader.load_fabric_agent_from_config(fabric_agent_config: dict) crewai.Crew¶
Load a complete Crew from configuration.
Args: fabric_agent_config: Configuration dict from get_fabric_agent_config().
Returns: Configured Crew instance ready to kickoff.
Raises: ImportError: If crewai is not installed.