Cheshire Cat

Cheshire Cat is a Python framework for learners, researchers and creatives building custom AI assistants. It supports hands-on experiments through a web chat interface and an API. It isn't recommended for production use because it's unstable and may have breaking changes or incomplete features.

• Multi-agent conversations: Use several agents in one chat, choose an agent for each message or let agents call each other.
• Custom assistant behavior: Add callable tools, document retrieval, memory and guardrails, with hooks to inspect or modify incoming tasks and outgoing replies.
• MCP connections: Give agents access to tools, prompts and resources supplied by Model Context Protocol servers.
• Chat and streaming: Manage multiple chats in the browser UI and receive generated text, tool calls and agent events through AG-UI streaming.
• Python extensions: Add agents and custom API endpoints through plugins, including model and authentication handlers, with role-based API access.

The framework runs on your own machine and supports Ollama and vLLM connections. It also connects to OpenAI, Anthropic, Gemini and OpenRouter; requests made through those cloud services go to the selected provider. The built-in UI runs locally in your browser, while the API lets developers use agents in their own applications. Cheshire Cat is open source under GPL-3.0.