pinboard-bookmarks-mcp-server
If you are the rightful owner of pinboard-bookmarks-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
The Pinboard MCP Server provides read-only access to Pinboard.in bookmarks for LLMs via the Model Context Protocol (MCP).
The Pinboard MCP Server is designed to facilitate interaction between large language models (LLMs) and Pinboard.in bookmarks. Built on the FastMCP 2.0 framework, this server allows LLMs to search, filter, and retrieve bookmark metadata at inference time. It offers a robust set of tools for bookmark interaction while adhering to Pinboard's API rate limits and implementing intelligent caching strategies. The server is equipped with four core MCP tools that enable efficient and effective access to bookmark data. It also features comprehensive testing and continuous integration validation to ensure reliability and performance. The server is easy to install via pip and can be quickly set up with a Pinboard API token. It is compatible with platforms like Claude Desktop, allowing seamless integration into existing workflows. The server's architecture leverages modern technologies such as FastAPI, Pydantic, and LRU caching to deliver fast and secure access to bookmark data.
Features
- Read-only access to Pinboard bookmarks
- Four MCP tools for bookmark interaction
- Smart caching with LRU cache and automatic invalidation
- Rate limiting to respect Pinboard's API guidelines
- Comprehensive testing with integration test harnesses
Tools
searchBookmarks
Search bookmarks by query string across titles, notes, and tags.
listRecentBookmarks
List bookmarks saved in the last N days.
listBookmarksByTags
List bookmarks filtered by tags with optional date range.
listTags
List all tags with their usage counts.