CrewAI

Developer Documentation

CrewAI provides a robust developer experience with multiple integration surfaces:

Installation

pip install crewai

Core Concepts

• Agents — Role-based entities with defined goals, backstories, and tool access.
• Tasks — Discrete units of work assigned to agents.
• Crews — Orchestrated groups of agents collaborating on a shared objective.
• Flows — Deterministic, controllable pipelines for complex multi-agent orchestration.

Code-First Example

from crewai import Agent, Task, Crew

researcher = Agent(role='Researcher', goal='Find key insights', ...)
writer = Agent(role='Writer', goal='Summarize findings', ...)
task = Task(description='Research and summarize AI trends', agent=researcher)
crew = Crew(agents=[researcher, writer], tasks=[task])
result = crew.kickoff()

Enterprise Features (Control Plane)

• Real-time tracing of every LLM call, tool call, and memory read
• RBAC with immutable audit trails and Enterprise IAM integration
• Human-in-the-loop intervention gates during execution
• Runtime hooks for PII redaction and policy enforcement at every LLM/tool boundary
• Multi-LLM model swapping at runtime for cost/performance optimization

Full API reference and CLI docs are available at docs.crewai.com.