Imagine an AI assistant that can browse a massive library of shared files for you. This tool acts as a digital bridge between AI models and YggTorrent, a popular file-sharing platform. It allows users to ask an AI to find specific content, such as software or media, and the AI can then search the site and provide download details directly within a chat conversation. It simplifies the process of finding and managing torrents by making them accessible through simple, natural language requests. Under the hood, this server provides a robust set of tools including specialized functions for searching torrents, retrieving detailed descriptions, and obtaining magnet links or .torrent files. It operates as a Python-based wrapper for the YggAPI, effectively translating complex web requests into a format that AI models can easily process. A key security feature is its local passkey injection system, which ensures that a user's sensitive credentials remain private and are only added to download links on the local machine rather than being transmitted to external servers. For developers and power users, this MCP server offers incredible flexibility by supporting multiple communication standards like stdio, SSE, and FastAPI. It can be seamlessly integrated into AI-powered development environments like Windsurf or Claude, enabling the LLM to act as an automated media curator or file manager. By providing a standardized interface for file-sharing data, it allows AI systems to autonomously handle the heavy lifting of content discovery and metadata retrieval, making it a powerful utility for building automated media pipelines or intelligent search agents.
Category: Files, Documents & PDFs
Tags: file-sharing, media, search, torrents, yggtorrent
bash pip install ygg-torrent-mcp Method 2: Using uv (Local Development) 1. Clone the repository: bash git clone https://github.com/philogicae/ygg-torrent-mcp.git cd ygg-torrent-mcp 2. Install dependencies: bash uv sync --locked Method 3: Using Docker 1. Clone the repository and navigate to the folder. 2. Build and run the container: bash docker compose up --build -d ---Mon compte -> PASSKEY field to find your key. 3. Set the environment variable YGG_PASSKEY or create a .env file in the project root: env YGG_PASSKEY=your_passkey_hereclaude_desktop_config.json): Using stdio (Recommended for most clients): json { "mcpServers": { "ygg-torrent-mcp": { "command": "uvx", "args": ["ygg-torrent-mcp"], "env": { "YGG_PASSKEY": "your_passkey_here" } } } } Using SSE transport (If running the server separately): json { "mcpServers": { "ygg-torrent-mcp": { "serverUrl": "http://127.0.0.1:8000/sse" } } } ---search_torrents: Search for torrents on the YggTorrent platform. * get_torrent_details: Retrieve detailed information for a specific torrent. * get_magnet_link: Retrieve the magnet link for a specific torrent (passkey is injected locally). * download_torrent_file: Retrieve the .torrent file for a specific torrent. * get_torrent_categories: Retrieve available torrent categories. ---leecher ratios, and upload dates. The AI can analyze the "details" of multiple torrents simultaneously to recommend the best option. Example: A user asks, "Find the best 4K HDR version of 'Dune: Part Two' with more than 100 seeders." The AI searches, compares the results provided by the MCP, and returns the magnet link for the highest-rated release.
.torrent file using the download_torrent_file tool, and save it directly into a NAS or Seedbox "watch" folder. Example: Through a chat interface like Windsurf, the user says, "I want to watch the latest episode of 'Top Chef France'. Find it on Ygg and save the torrent file to my `home/media/watch` folder." The AI handles the search, passkey injection, and file movement in one step.
search_torrents tool to perform this check.Part of MCP Servers