chainlit-mcp-sql-demo

This repository demonstrates a minimal Chainlit UI integrated with an agentic MCP (Model-Centric Programming) framework that invokes a SQL agent powered by an Anthropic LLM. The system supports natural language queries that are translated into SQL and executed, with results displayed through the Chainlit frontend.

---

🚀 Project Goal

To build a working prototype that combines:

• Chainlit UI: An open-source framework for building LLM-powered user interfaces.
• MCP agentic flow: Agentic infrastructure to orchestrate tool invocation and memory handling.
• Anthropic LLM: Claude model for natural language understanding.
• Text2SQL agent: Converts natural queries to SQL statements for query execution.

Framework	Key Features	Strengths	Ideal Use Cases	References
Anthropic Claude and MCP Agentic Framework​	Natural language to SQL, agent orchestration, Chainlit UI, LLM integration, API-ready, conversational, modular, Pythonic, dotenv config, SQLAlchemy support, interactive chat	Seamless natural language processing, modular architecture, multi-agent flow, easy UI setup, scalable logic, developer-friendly, portable, and well-documented	LLM UI demos, SQL automation, teaching assistants	chainlit-mcp-sql-demo YouTube

---

📦 Installation Instructions

Note:

• On macOS, use Terminal directly.
• On Windows, open WSL by pressing the Windows key and launching a Linux shell.

Step-by-step setup

# 1. Install uv if you haven't
pip install uv

# 2. Initialize a new project environment
uv init proj
cd proj

# 3. Add required packages
uv add anthropic chainlit mcp pandas python-dotenv sqlalchemy tabulate

# 4. Activate virtual environment
source .venv/bin/activate

# 5. Launch the Chainlit app
chainlit run app.py

---

👤 Author

Yiqiao Yin Developer and investor with a focus on generative AI, large language models, and intelligent applications.

---

🧠 Notes

• Ensure you have a .env file configured with your API keys (e.g., for Anthropic).
• This demo uses a basic SQLite backend for executing SQL queries. You can extend it to work with other databases.
• Contributions and improvements welcome!

---

📄 License

MIT License — feel free to use, modify, and share.