Official MCP server for GitHub: repos, issues, PRs, actions and code search for AI assistants.
Open Source Development
MCP Servers
Model Context Protocol servers — connect AI assistants to tools, APIs, databases and services.
6 listings in this category
Microsoft's MCP server for browser automation — let AI agents navigate, click and extract from real web pages.
Open Source Browser Automation
Up-to-date, version-specific library documentation injected straight into your AI coding assistant.
Free Documentation
Reference MCP server for secure, sandboxed local file system access — read, write and search files.
Open Source System
Query Postgres databases from AI assistants: schema inspection and read-only SQL over MCP.
Open Source Databases
Bring production error context into your AI assistant: issues, stack traces and debugging over MCP.
Free Observability
Nothing in this subcategory yet — submit a tool!
About MCP Servers
MCP Servers are tools and services that implement the Model Context Protocol — an open standard that lets AI assistants connect to external systems without custom integration code. An MCP server might expose access to a database, a SaaS API, a local file system, a code execution environment, or a specialized data source. When a capable client (like Claude Desktop, Cursor, or Cline) connects to an MCP server, the AI gains the ability to read and write to that system during a conversation. The directory includes official servers published by major platforms (GitHub, Notion, Figma, Slack) as well as community-built servers for hundreds of specialized tools. When evaluating MCP servers, look at authentication methods, what operations are exposed (read vs. read/write), rate limit handling, and whether the server is actively maintained. The ecosystem is growing rapidly — new servers are published daily.