Linkedin_Mcp_Server

# LinkedIn Profile Scraper MCP Server This MCP server uses the Fresh LinkedIn Profile Data API to fetch LinkedIn profile information. It is implemented as a model context protocol (MCP) server and exposes a single tool, `get_profile`, which accepts a LinkedIn profile URL and returns the profile data in JSON format. ## Features - **Fetch Profile Data:** Retrieves LinkedIn profile information including skills and other settings (with most additional details disabled). - **Asynchronous HTTP Requests:** Uses `httpx` for non-blocking API calls. - **Environment-based Configuration:** Reads the `RAPIDAPI_KEY` from your environment variables using `dotenv`. ## Prerequisites - **Python 3.7+** – Ensure you are using Python version 3.7 or higher. - **MCP Framework:** Make sure the MCP framework is installed. - **Required Libraries:** Install `httpx`, `python-dotenv`, and other dependencies. - **RAPIDAPI_KEY:** Obtain an API key from [RapidAPI](https://rapidapi.com/) and add it to a `.env` file in your project directory (or set it in your environment). ## Installation 1. **Clone the Repository:** ```bash git clone https://github.com/codingaslu/Linkedin_Mcp_Server cd Linkedin_Mcp_Server ``` 2. **Install Dependencies:** ```bash uv add mcp[cli] httpx requests ``` 3. **Set Up Environment Variables:** Create a `.env` file in the project directory with the following content: ```ini RAPIDAPI_KEY=your_rapidapi_key_here ``` ## Running the Server To run the MCP server, execute: ```bash uv run linkedin.py ``` The server will start and listen for incoming requests via standard I/O. ## MCP Client Configuration To connect your MCP client to this server, add the following configuration to your `config.json`. Adjust the paths as necessary for your environment: ```json { "mcpServers": { "linkedin_profile_scraper": { "command": "C:/Users/aiany/.local/bin/uv", "args": [ "--directory", "C:/Users/aiany/OneDrive/Desktop/linkedin-mcp/project", "run", "linkedin.py" ] } } } ``` ## Code Overview - **Environment Setup:** The server uses `dotenv` to load the `RAPIDAPI_KEY` required to authenticate with the Fresh LinkedIn Profile Data API. - **API Call:** The asynchronous function `get_linkedin_data` makes a GET request to the API with specified query parameters. - **MCP Tool:** The `get_profile` tool wraps the API call and returns formatted JSON data, or an error message if the call fails. - **Server Execution:** The MCP server is run with the `stdio` transport. ## Troubleshooting - **Missing RAPIDAPI_KEY:** If the key is not set, the server will raise a `ValueError`. Make sure the key is added to your `.env` file or set in your environment. - **API Errors:** If the API request fails, the tool will return a message indicating that the profile data could not be fetched. ## License This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for more details.