Welcome to motleycrew’s documentation!

Contents:

• Home
• Installation
  • Installation using pip
  • Installation from source
• Usage
  • Quickstart
    • Agents and tools
    • Crew and tasks
    • Knowledge graph backend and custom tasks
  • Key Concepts and API
    • Crew and knowledge graph
    • Tasks, task units, and workers
    • Task hierarchy
    • How the crew handles tasks
    • Asynchronous execution
  • Agents
    • ReAct tool calling agent
    • Using agents from other frameworks
  • 2025 update on CrewAI agents
    • Creating your own agent
  • Interacting with the knowledge graph
  • Choosing LLMs
    • Providing an LLM to an agent
    • Default LLM
    • Using custom LLMs
  • Caching and observability
    • Lunary and self-hosting
  • Key-value store for manipulating objects
• Examples
  • Blog with images
  • Multi-step research agent
  • Using an output handler for validating agent’s output
  • Advanced output handling
  • Customer support chatbot with Ray Serve
    • Key Components
    • Implementation Details
    • Running the Example
  • Streaming agent output
    • Streaming agent output to a web client using WebSockets
  • Event-driven orchestration for AI systems: new ways and old
    • Using context
    • Constructing simple workflows in a concise manner using FaustWorkflow
  • AutoGen-related Examples
    • Math via python code with a single agent
    • Using AutoGen with motleycrew
• API
  • motleycrew
    • motleycrew.agents
    • motleycrew.applications
    • motleycrew.common
    • motleycrew.crew
    • motleycrew.storage
    • motleycrew.tasks
    • motleycrew.tools
    • motleycrew.tracking

Indices and tables

• Index

• Module Index

• Search Page