LinkedIn-Posts-Hunter

LinkedIn Posts Hunter MCP Server exposes six tools that help you automate, search, and manage LinkedIn job posts discovered through LinkedIn posts rather than traditional job boards. Use auth to login and persist your session, search_posts to scrape new posts by keywords, and manage_posts to read, update, or delete entries in your local SQLite database. The viewer_filters tool lets you programmatically adjust the React dashboard filters, and start_viewer / stop_viewer control the local React-based dashboard UI. You can mix natural language commands with direct tool invocations from your AI assistant, with results being reflected in the local UI and database.

Prerequisites:\n- Node.js 18+ installed on your machine\n- npm comes with Node.js\n- A LinkedIn account to authenticate\n- Cursor IDE or Claude Desktop MCP clients\n\nMethod A: Using mcp.json Configuration (Recommended)\n1. Install the MCP server globally:\nbash\nnpm install -g linkedin-posts-hunter-mcp\n\n2. Add to your MCP configuration. For Cursor IDE, edit mcp.json (path varies by platform) and add:\njson\n{\n "mcpServers": {\n "linkedin-posts-hunter-mcp": {\n "command": "linkedin-posts-hunter-mcp"\n }\n }\n}\n\nFor Claude Desktop, edit claude_desktop_config.json and add the same server entry.\njson\n{\n "mcpServers": {\n "linkedin-posts-hunter-mcp": {\n "command": "linkedin-posts-hunter-mcp"\n }\n }\n}\n\n3. Restart your MCP client (Cursor or Claude Desktop).\n\nMethod B: Local Development Setup (for contributors)\n1. Clone the repository and install dependencies:\nbash\ngit clone https://github.com/kevin-weitgenant/LinkedIn-Posts-Hunter-MCP-Server.git\ncd LinkedIn-Posts-Hunter-MCP-Server\nnpm install\nnpm run build\n\n2. To run locally without global install, start via Node with the built entry point (path may vary after build):\nbash\nnode build/index.js\n

Notes & tips:\n- The authentication flow uses Playwright to log into LinkedIn in a browser window; cookies are stored locally for persistent sessions. Ensure you have a secure, local environment for storing credentials.\n- All scraped posts are stored in a local SQLite database on your machine. Your data does not leave your computer unless you export it.\n- The React dashboard runs as a separate UI that reads from the local database; use viewer_filters to programmatically adjust UI filters or manage_posts to modify the dataset.\n- If you run into issues with the viewer, ensure your local firewall or antivirus isn’t blocking the local server used by the dashboard.\n- Common commands: "Authenticate my LinkedIn account", "Search LinkedIn for 'Python developer' jobs", "Show me posts I haven't applied to yet".