mcp_demo

# KPI Dashboard MCP Server An intelligent KPI Dashboard server built on FastAPI and FastMCP. It supports local data analysis and agent interaction, suitable for enterprise operations, business monitoring, intelligent analysis, and other scenarios. --- ## 🌟 Project Highlights - ✅ Lightweight and high-performance HTTP interface built with FastAPI - 🤖 Integrated FastMCP Server, supports intelligent dialogue with MCP clients (such as Cherry Studio) - 📊 Built-in data analysis module, no external database or CSV files required - 🔧 Supports local development debugging and VS Code plugin usage --- ## 📁 Project Structure ``` kpi-dashboard-mcp-server/ │ ├── api.py # Provides APIs related to data analysis ├── helper.py # Data analysis utility functions (e.g., KPI calculation) ├── main.py # Starts the MCP Server service ├── data.py # Built-in simulated data, replaces CSV loading ├── requirements.txt # Required Python dependencies └── README.md # Project documentation ``` --- ## 🚀 Quick Start ### 1. Clone the Repository ```bash git clone https://github.com/zhangting-hit/mcp_demo.git cd mcp_demo ``` ### 2. Create a Virtual Environment (Optional) ```bash python -m venv venv # macOS/Linux: source venv/bin/activate # Windows: venv\Scripts\activate ``` ### 3. Install Dependencies ```bash pip install -r requirements.txt ``` --- ## ▶️ Start the Service ### Start the API Service (default port `8000`) ```bash python api.py ``` Access the documentation at: * [http://127.0.0.1:8000/docs](http://127.0.0.1:8000/docs) ### Start the MCP Server Service (default port `8080`) ```bash python main.py ``` After starting, the output will be: ``` FastMCP Server is running at http://localhost:8080 ``` > Example Screenshots >  >  --- ## 🔗 Connect MCP Server in Cherry Studio 1. Open Cherry Studio 2. Go to Settings → Add MCP Service Configuration 3. Use the following content: ```json { "name": "KPI Analysis Agent", "command": "D:\\Anaconda\\envs\\mcp\\python.exe", "args": [ "D:/Users/Desktop/mcp_demo/main.py" ], "env": { "PYTHONUNBUFFERED": "1" } } ``` Once connected successfully, you can converse with the Agent through Cherry. >  --- ## 📡 API Interface Description ### `GET /kpi/summary` Returns the KPI summary results. #### Example Response: ```json { "total_sales": 150000, "conversion_rate": 0.035, "active_users": 3421 } ``` --- ## 💡 Example: Adding New KPI Analysis Logic Add the following in `helper.py`: ```python def calculate_average_order_value(sales_data): return sum(s["amount"] for s in sales_data) / len(sales_data) ``` Then use it in the `/kpi/summary` endpoint in `api.py`. --- ## 📚 Tech Stack * [FastAPI](https://fastapi.tiangolo.com/) * [FastMCP](https://github.com/zhplus/fastmcp) * Python 3.8+ --- ## 📌 Frequently Asked Questions ### Q: Accessing `http://127.0.0.1:8000/` shows 404? A: Accessing the root path directly will return 404. Please use the following paths: * [http://127.0.0.1:8000/kpi/summary](http://127.0.0.1:8000/kpi/summary) * [http://127.0.0.1:8000/docs](http://127.0.0.1:8000/docs) --- ## 🧑‍💻 Developer Suggestions * You can replace `data.py` with a database or external API data source * MCP Server supports extending context memory, chart output, and multi-turn Q&A * Can be used to build embedded knowledge bases, BI platforms, customer operation analysis systems, etc. --- ## 📄 License MIT License