Model Context Protocol (MCP) servers act as bridges between AI models and external services, enabling AI to interact with tools like Google Drive, Figma, databases, and browsers. A curated list of 100 popular MCP servers has been compiled, collectively gathering 3.1 million stars on GitHub.
These tools allow AI to perform specific tasks such as searching for documents, managing code repositories, and automating browser actions. Users can install these servers via terminal commands and integrate them into platforms like Claude, Cursor, or Codex by providing necessary API keys or account credentials.
The collection covers various categories, including search and data retrieval, software development, and file management. For instance, specific servers enable AI to query databases like PostgreSQL or SQLite, analyze sales data in Metabase, or manage tasks in Jira and Confluence.
Development-focused servers facilitate code analysis, version control via Git, and error tracking through Sentry. Additionally, browser-based servers allow AI to navigate websites, fill out forms, and perform repetitive tasks, while file-management tools enable reading and editing local files or notes in applications like Obsidian.
While many MCP servers are open-source and free to run locally, some require paid subscriptions or specific API access to the services they connect with. An expanded version of the list is also available, featuring double the number of tools compared to the primary selection.
