ayn 0.0.2

Agents You Need -- Tools to search and access the AI agents you need

AYN - Agents You Need

The meta-framework for AI agent interoperability

Think "pip + Docker Hub + Postman" for AI agents.

🎯 Vision

AYN solves the agent discovery and integration problem by providing:

1. Unified Search across GitHub, HuggingFace, and awesome lists
2. Standard Controller Interface for invoking any agent
3. HTTP/REST Export via FastAPI and LangServe
4. MCP Support for Model Context Protocol integration
5. Custom Actions for ChatGPT and Claude

🚀 Quick Start

from ayn import AgentRegistry, BaseAgentController, export_agent_full_stack

# 1. Search for agents
registry = AgentRegistry()
results = registry.search('data preparation')

# 2. Create controller
controller = BaseAgentController.from_metadata(results[0])

# 3. Use the agent
result = controller.invoke({'data': [1, 2, 3]})

# 4. Export with all integrations
artifacts = export_agent_full_stack(
    controller,
    results[0],
    export_fastapi=True,
    export_mcp=True,
    export_chatgpt=True
)

📦 Installation

# Basic installation
pip install ayn

# With FastAPI support
pip install ayn[fastapi]

# With LangServe support
pip install ayn[langserve]

# Full installation
pip install ayn[all]

📚 Documentation

• Getting Started - Quick introduction and setup
• README - Complete documentation
• API Reference - Detailed API documentation
• Quick Reference - Common patterns and snippets
• Implementation Roadmap - Future development plans

🏗️ Architecture

┌─────────────────────────────────────────────────────────────┐
│                      AYN Framework                          │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐    │
│  │   Search     │  │  Controller  │  │    Export    │    │
│  │   Registry   │  │  Interface   │  │    Layer     │    │
│  └──────────────┘  └──────────────┘  └──────────────┘    │
│         │                  │                  │            │
│  ┌──────▼──────┐  ┌────────▼────────┐  ┌─────▼─────┐    │
│  │  GitHub     │  │  BaseAgent      │  │  FastAPI  │    │
│  │  HuggingFace│  │  Controller     │  │  LangServe│    │
│  │  AwesomeLists│ └─────────────────┘  └───────────┘    │
│  └─────────────┘                                          │
│                                                             │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐    │
│  │  MCP Layer   │  │   Actions    │  │  Frameworks  │    │
│  │  Tools       │  │   ChatGPT    │  │  CrewAI      │    │
│  │  Resources   │  │   Claude     │  │  LangChain   │    │
│  │  Prompts     │  │   Gemini     │  │  AutoGen     │    │
│  └──────────────┘  └──────────────┘  └──────────────┘    │
└─────────────────────────────────────────────────────────────┘

🎓 Examples

See the examples/ directory for complete working examples:

• basic_usage.py - Basic registry and controller usage
• fastapi_deployment.py - Deploy as HTTP API
• mcp_integration.py - MCP protocol integration
• tutorial.py - Comprehensive tutorial

🗺️ Roadmap

Phase 1: Core ✅ (Complete)

• ✅ Agent metadata and registry
• ✅ Standard controller interface
• ✅ FastAPI export
• ✅ MCP server support
• ✅ ChatGPT/Claude actions

Phase 2: Search (In Progress)

• ⏳ GitHub API integration
• ⏳ HuggingFace search
• ⏳ Awesome list parsing
• ⏳ Caching and indexing

Phase 3: Installation (TODO)

• ⬜ Dependency management (uv/pipx)
• ⬜ Docker support
• ⬜ Environment isolation
• ⬜ Version management

Phase 4: Frameworks (TODO)

• ⬜ CrewAI integration
• ⬜ LangChain/LangGraph integration
• ⬜ AutoGen integration
• ⬜ SmolAgents integration

See AYN_IMPLEMENTATION_ROADMAP.md for detailed plans.

🤝 Contributing

Contributions welcome! Key areas:

1. Searchers: Implement real API integrations
2. Controllers: Add framework-specific controllers
3. Export: Additional export formats
4. Tools: MCP tools and resources
5. Documentation: Examples and guides

📄 License

MIT License - see LICENSE file

🙏 Acknowledgments

Built on top of:

• FastAPI for HTTP APIs
• LangServe for LangChain integration
• Model Context Protocol (Anthropic)

---

AYN - Making AI agents discoverable, interoperable, and deployable. 🚀