The open directory of Model Context Protocol servers — browse, search, and integrate tools into your AI workflows.
The complete platform for discovering, evaluating, and deploying MCP servers into production AI workflows.
Browse a curated directory of official and community-built MCP servers across databases, developer tools, cloud services, and more. Filter by language, transport type, and category.
Every server comes with ready-to-use install commands and JSON config examples. Copy and paste directly into your Claude Desktop or AI client config.
Dead and unreachable servers are automatically removed from the directory. The install configs you copy here point to servers that are actually up and running.
Each server entry lists every tool it exposes — from read_file to kubernetes_deploy — so you know exactly what capabilities you're adding to your AI. Complete with transport type, version history, and configuration examples.
Find the right integration for every part of your stack.
Hand-picked, production-ready integrations
Real-time infrastructure monitoring with metrics, logs, alerts, and ML-based anomaly detection.
Draw and visually collaborate with your agents on tldraw's canvas.
MCP server for Chrome DevTools
Validate oh-my-posh configurations and segment snippets against the official schema.
Search and discover Agent Skills from the skills.sh registry. Powered by HAPI MCP server.
Publish and discover MCP servers via the official MCP Registry. Powered by HAPI MCP server.
Fresh releases and recent updates
Everything you need to know about MCP servers and MCP Find.
MCP Find is an open-source directory of Model Context Protocol servers. It helps developers discover, evaluate, and install MCP servers into AI clients like Claude Desktop, Cursor, VS Code, and Windsurf.
MCP is an open standard created by Anthropic that lets AI assistants connect to external tools and data sources. Think of it like a USB port for AI — any MCP-compatible client can use any MCP server, regardless of who built it.
Find a server on MCP Find, copy the JSON config snippet from its listing page, and paste it into your AI client's configuration file. Most servers work with a single command or config block — no complex setup required.
The vast majority of MCP servers listed here are open source and free. Some connect to paid third-party services (like cloud providers or SaaS APIs), but the MCP server itself is typically free.
Claude Desktop, Claude Code, Cursor, VS Code (via Copilot), Windsurf, and a growing number of AI tools support MCP. Any client that implements the protocol can connect to the servers listed here.
Click "Submit a Server" in the navigation bar and fill out the form with your server's package name and GitHub URL. Submissions are reviewed and typically appear in the directory within a few days.
Guides, tutorials, and insights on MCP
Datadog has an official MCP server that lets Claude and Cursor query metrics, read logs, check monitors, and surface dashboards. This guide covers installation, authentication, the full tool surface, and how to scope your API key for safe read-only agent access to production observability data.
MCPFind indexes 14,899 MCP servers across 21 categories, including tools for location data and mapping. This guide shows how to set up the Google Maps MCP server, which APIs it supports, and how to build location-aware AI workflows using geocoding, routing, and places data.
The OpenAI Agents SDK and MCP are not competing standards. OpenAI added native MCP support to the Responses API in May 2025. This guide shows when to use SDK built-in tools, when MCP server portability matters, and how to wire both together in the same agent.
Share it with the community. MCP Find makes it easy for developers to discover your integration and start using it in minutes.