A framework for building custom AI agents as microservices with API-first approach, conversational interfaces, plugin extensibility, built-in RAG capabilities, and multiuser support with granular permissions.
One Minute Overview#
Cheshire Cat AI is a framework for building custom AI agents as microservices, designed for developers who need to integrate conversational AI capabilities into their applications. It provides easy-to-use APIs and WebSocket communication, built-in Retrieval-Augmented Generation (RAG) functionality, and extensibility through a plugin system.
Core Value: Enables developers to quickly build and customize AI assistants without having to handle complex AI infrastructure from scratch.
Quickstart#
Installation Difficulty: Low - Only requires Docker
docker run --rm -it -p 1865:80 ghcr.io/cheshire-cat-ai/core:latest
Is this suitable for my needs?
- ✅ Projects needing to integrate AI assistants into existing applications
- ✅ Building intelligent systems with conversational forms and tool calling
- ❌ Projects requiring highly customized AI model architectures
Core Capabilities#
1. Plugin System#
- Supports three types of plugins: hooks, tools, and conversational forms for flexible AI agent extension Actual Value: Add new functionality without modifying core code, such as custom AI personas, specialized tools, and structured conversation flows
2. Built-in RAG Functionality#
- Based on Qdrant vector database, supports Retrieval-Augmented Generation Actual Value: AI agents can provide accurate responses based on private knowledge bases, reducing hallucination issues
3. Multi-user with Permissions#
- Supports multi-user environments with granular permissions, compatible with any identity provider Actual Value: Securely build AI applications for enterprises or teams with data isolation and access control
4. Multi-model Support#
- Supports any language model through LangChain Actual Value: Not limited to specific AI providers, can flexibly choose the most suitable model based on needs
Tech Stack & Integration#
Development Language: Python Key Dependencies: LangChain, Qdrant, Pydantic, FastAPI (WebSocket/REST API) Integration Method: REST API / WebSocket / SDK (via plugins)
Maintenance Status#
- Development Activity: Actively developed with a dedicated roadmap and contribution guidelines
- Recent Updates: Project continuously updated with detailed version planning and feature iteration roadmap
- Community Response: Active Discord community and extensive documentation resources
Commercial & Licensing#
License: GPL-3.0
- ✅ Commercial: Allowed, but requires attribution and source code disclosure
- ✅ Modification: Allowed
- ⚠️ Restrictions: Follows GPL protocol, derived works must be open-sourced
Documentation & Learning Resources#
- Documentation Quality: Comprehensive
- Official Docs: cheshire-cat-ai.github.io/core/
- Example Code: Provides plugin development examples including hooks, tools, and conversational forms