FastAgentic - The Deployment Layer for AI Agents

FastAgentic is a production deployment framework for AI agents. It provides a simple
way to deploy agents built with any framework (PydanticAI, LangChain, LangGraph,
CrewAI, DSPy, LlamaIndex, and more) and exposes them through REST APIs, MCP servers,
and A2A interfaces.

Key Features:
- Universal adapter system for any agent framework
- Automatic REST API generation with OpenAPI docs
- MCP (Model Context Protocol) server support
- A2A (Agent-to-Agent) communication
- Built-in authentication with OIDC
- Durable execution with checkpointing
- Real-time streaming via SSE
- Rate limiting and policy enforcement
- Observability with OpenTelemetry

Getting Started:
1. Install FastAgentic: pip install fastagentic
2. Create an App with your agent
3. Use @agent_endpoint to expose your agent
4. Run with: fastagentic run

Example:
```python
from fastagentic import App, agent_endpoint
from fastagentic.adapters.llamaindex import LlamaIndexAdapter

app = App(title="My Agent")

@agent_endpoint(path="/query", runnable=LlamaIndexAdapter(query_engine))
async def query(question: str) -> str:
    pass
```

For more information, visit the documentation at https://github.com/neul-labs/fastagentic
