Implementation:CrewAIInc CrewAI Crew Constructor

Knowledge Sources	CrewAI CrewAI Docs
Domains	Multi_Agent_Systems, Orchestration
Last Updated	2026-02-11 00:00 GMT

Overview

Concrete class for assembling agents, tasks, and configuration into an executable crew provided by the CrewAI framework.

Description

The Crew class is the central orchestration container. It accepts lists of agents and tasks, a process type (sequential or hierarchical), and optional configuration for memory, caching, verbose logging, embedders, knowledge sources, callbacks, planning, and streaming. Pydantic validators ensure consistency (e.g., hierarchical process requires manager_agent or manager_llm).

Usage

Import and instantiate Crew after creating Agent and Task instances. Pass the agents, tasks, and desired process type. The resulting Crew instance is ready for kickoff().

Code Reference

Source Location

• Repository: crewAI
• File: lib/crewai/src/crewai/crew.py
• Lines: L133-326

Signature

class Crew(FlowTrackable, BaseModel):
    """Represents a group of agents and their tasks."""

    name: str | None = Field(default="crew")
    agents: list[BaseAgent] = Field(default_factory=list)
    tasks: list[Task] = Field(default_factory=list)
    process: Process = Field(default=Process.sequential)
    verbose: bool = Field(default=False)
    memory: bool = Field(default=False)
    cache: bool = Field(default=True)
    max_rpm: int | None = Field(default=None)
    embedder: EmbedderConfig | None = Field(default=None)
    manager_llm: str | BaseLLM | Any | None = Field(default=None)
    manager_agent: BaseAgent | None = Field(default=None)
    function_calling_llm: str | LLM | Any | None = Field(default=None)
    planning: bool | None = Field(default=False)
    planning_llm: str | BaseLLM | Any | None = Field(default=None)
    knowledge_sources: list[BaseKnowledgeSource] | None = Field(default=None)
    stream: bool = Field(default=False)
    output_log_file: bool | str | None = Field(default=None)

Import

from crewai import Crew, Process

I/O Contract

Inputs

Name	Type	Required	Description
agents	list[BaseAgent]	Yes	List of agents in the crew
tasks	list[Task]	Yes	Ordered list of tasks
process	Process	No	Execution process type (default: Process.sequential)
verbose	bool	No	Enable verbose logging (default: False)
memory	bool	No	Enable memory subsystem (default: False)
cache	bool	No	Enable tool result caching (default: True)
embedder	None	No	Embedding configuration for knowledge/memory

Outputs

Name	Type	Description
Crew instance	Crew	Assembled crew ready for kickoff()

Usage Examples

Sequential Crew

from crewai import Crew, Agent, Task, Process

crew = Crew(
    agents=[researcher, writer],
    tasks=[research_task, writing_task],
    process=Process.sequential,
    verbose=True,
    memory=True,
)

Related Pages

Implements Principle

• Principle:CrewAIInc_CrewAI_Crew_Assembly
• Environment:CrewAIInc_CrewAI_Python_Runtime_Environment