MCP Encyclopedia

microsoft/markitdown Details →

Featured Python

178k

An official Microsoft document-conversion tool that turns Word, PDF, Excel and PPT into Markdown in one click, so AI can read your documents directly.

punkpeye/awesome-mcp-servers Details →

Featured

93.7k

The most authoritative collection of MCP servers on GitHub: hundreds of servers organized by use case. The first place to look when hunting for MCP tools.

Panniantong/Agent-Reach Details →

Featured Python

77.4k

Let your AI agent see the entire internet: read and search content from Twitter, Reddit, YouTube, GitHub and more. An essential tool for information-seeking agents.

A context-optimization layer for LLM applications that automatically manages the context window — dramatically improving agent performance in long conversations and large codebases.

Indexes an entire codebase into a persistent knowledge graph: supports 35 languages with millisecond queries, roughly 99% faster than grep. Long-term memory for coding AI agents.

microsoft/playwright-mcp Details →

Featured TypeScript

36.7k

An official Microsoft MCP server for browser automation: lets AI agents control a browser to click, fill forms, screenshot and scrape. The go-to for web automation.

github/github-mcp-server Details →

Featured Go

32.7k

The official GitHub MCP server: lets AI agents directly manage repositories, PRs and issues. A core component of developer automation workflows.

AIHawk, an automated job-application assistant: a browser-based AI agent that browses job listings, customizes your resume and applies for you. A job-hunting automation wonder.

Lets AI command Blender through natural language for 3D modeling, scene creation and editing. A flagship project for creating 3D assets with prompts.

The official MCP Python SDK: quickly build MCP servers and clients in Python. Officially maintained with thorough docs — the standard starting point for developers.

A lightweight cross-platform database client at just 20MB, with built-in MCP support that lets AI agents query and manage your databases directly.

The official MCP TypeScript SDK: the best way to build MCP servers and clients in the JS/TS ecosystem, seamless with the Node stack.

A Chrome-operating MCP: lets AI drive the browser to click, fill forms and screenshot — a lightweight way to automate web operations.

A big curated collection of MCPs for Chinese users: a search-friendly list of MCP servers, the first stop for developers in China.

A broad collection of AI MCP servers: hundreds of servers organized by use case, actively community-maintained. A complementary entry point for finding MCPs.

Connects to 3,000+ pre-built cloud scraping tools (Actors) on the Apify platform, covering websites, e-commerce, social media, search engines and more.

The official MCP Go SDK: write high-performance MCP servers and clients in Go. Great for building production-grade, high-concurrency MCP tools.

The Home Assistant smart-home MCP: lets AI control devices in your home and query their status — build your own AI smart-home hub.

The official MCP C# SDK: the standard way for .NET developers to integrate MCP, covering both servers and clients. Suited to enterprise projects.

One of the earliest MCP server collections: broad coverage with indexes of common browser, database, filesystem and developer-tool servers.

The official MCP Java SDK: build MCP servers and clients in the Java ecosystem, friendly to Spring and other frameworks. A top pick for back-end teams.

A collection of official Microsoft MCP deployment solutions: centralizes MCP integration for multiple Microsoft products, letting agents interop with the Microsoft ecosystem.

A set of MCP adapters and tools: provides MCP integration adapters for common applications, lowering the barrier to building servers and ecosystem integration yourself.

An MCP app store: packages MCP servers and clients into one-click installable products. A distribution platform for the MCP ecosystem.

A curated collection of Claude plugins: gathers plugins, extensions and skills around Claude Code / Copilot to enhance everyday development.

A document-parsing MCP: parses PDF, Word and other document types into structured content — a preprocessing tool before feeding content to AI.

MiniMax's official MCP: a one-stop server that brings image generation, text-to-video, speech and other multimodal capabilities into your AI tools.

The Qdrant vector-database MCP: lets AI operate a knowledge base directly for semantic retrieval — the standard vector storage for RAG applications.

An AI accessibility-testing browser: lets agents automatically run accessibility (a11y) checks on web pages, helping with QA and compliance.

A collection of DevOps-oriented MCP servers: Kubernetes, IaC, monitoring and CI/CD servers, curated for DevOps engineers.