MCP Router Server

A FastAPI-based MCP-compliant server that routes requests to multiple LLM providers (OpenAI, LM Studio, OpenRouter, Ollama, Claude, Azure, etc.) via a unified REST API.

Features

Switch LLM providers by editing .env
/ask endpoint for context-aware chat requests
Health check at /health
Provider abstraction for OpenAI, LM Studio, OpenRouter, Ollama, Anthropic Claude, Azure Foundry AI
Easy deployment and configuration

Quickstart

Install dependencies
(Python 3.8+ required, only pure Python packages supported in this environment)
```
pip install -r requirements.txt
```

Copy and edit environment variables

cp .env.example .env
# Edit .env to set your provider keys and options

Run the server
```
uvicorn app.main:app --reload
```
Test health check
```
curl http://localhost:8000/health
```

Test /ask endpoint

curl -X POST http://localhost:8000/ask \

-H "Content-Type: application/json"
-d '{ "identity": { "user_id": "demo" }, "memory": { "history": [ { "role": "system", "content": "You are a helpful assistant that only answers in haikus." }, { "role": "user", "content": "who are you?" } ] }, "tools": [], "docs": [], "extra": {} }'


### Anthropic

curl -X POST http://localhost:8000/ask
-H "Content-Type: application/json"
-d '{ "identity": { "user_id": "demo" }, "memory": { "history": [ { "role": "user", "content": "You are a helpful assistant that only answers in haikus.\n\nwho are you?" } ] }, "tools": [], "docs": [], "extra": {} }'


## Environment Variables

See `.env.example` for all supported variables.

## Project Structure

app/ init.py main.py router.py .env.example requirements.txt .gitignore LICENSE README.md


## License

MIT License