CrewAI

CrewAI is a lean, fast multi-agent orchestration framework built on a role-playing metaphor. You define a "crew" of agents, each with a distinct role, backstory, goal, and toolset. CrewAI handles task delegation, inter-agent communication, and result synthesis automatically.

Key features

• Role-based agents — give each agent a persona, goal, and backstory for more focused behavior
• Sequential and hierarchical processes — choose how tasks flow between agents
• Tool use — agents can use search, code execution, file I/O, and custom tools
• Memory — short-term, long-term, and entity memory across tasks
• Async execution — run agents concurrently for faster pipelines
• Model-agnostic — works with OpenAI, Anthropic, Groq, Ollama, and more

Quick start

npx ai-supply add crewai-multi-agent

# Or install directly
pip install crewai crewai-tools

from crewai import Agent, Task, Crew

researcher = Agent(
    role="Research Analyst",
    goal="Find key facts about a topic",
    backstory="You are a meticulous researcher.",
    verbose=True
)

writer = Agent(
    role="Content Writer",
    goal="Write a concise summary from research findings",
    backstory="You craft clear, engaging summaries.",
    verbose=True
)

task1 = Task(description="Research the history of the MCP protocol", agent=researcher)
task2 = Task(description="Write a 3-paragraph summary of the research", agent=writer)

crew = Crew(agents=[researcher, writer], tasks=[task1, task2])
result = crew.kickoff()
print(result)

Curated mirror of the open-source CrewAI project (MIT). Install upstream from the repository.