Open Source MCP Projects
Discover 141 open source MCP repositories from GitHub, each with an analysis of what it does, key features, use cases and alternatives. MCP projects here are most often combined with AI Agents, TypeScript and Python. Last updated October 3, 2026.
141 repositories · updated October 3, 2026
codegraph: Give Coding Agents a Local Code Knowledge Graph
CodeGraph builds and maintains a local graph of code structure so AI coding agents can retrieve relevant source, call paths, and change impact without exploring files one by one. It connects to multiple agents through MCP and supports many languages.
ai-engineering-from-scratch: Learn AI by Building It
A free, MIT-licensed curriculum that teaches AI engineering through runnable lessons and hands-on artifacts. It covers foundations through LLMs, agents, protocols, and production, with paths for beginners and experienced developers.
apm: Manage Dependencies for AI Agent Setups
APM manages AI-agent instructions, skills, prompts, plugins, and MCP servers through a project manifest and lockfile. It helps teams reproduce agent configurations across supported coding clients and apply installation security and policy controls.
claude-mem: Preserve Agent Context Across Sessions
Claude-Mem captures and summarizes agent activity, then retrieves relevant project history in later sessions. It is for developers who want coding agents to retain useful context instead of starting from scratch each time.
graphify: Turn Codebases into Queryable Knowledge Graphs
Graphify builds a queryable knowledge graph from code, documentation, schemas, and other project files. It helps developers and AI coding assistants explore relationships, trace paths, and understand codebases without relying on a vector store.
agentic-inbox: Manage Email with a Self-Hosted AI Assistant
Agentic Inbox is a self-hosted email client that runs on Cloudflare Workers and adds an AI agent for inbox tasks. It suits people already using Cloudflare who want email and AI tools under their own account, with Cloudflare Access as the shared security boundary.
claude-for-legal: Legal Workflow Plugins and Agents
Anthropic’s claude-for-legal provides Claude plugins, skills, and managed-agent cookbooks for legal teams, clinics, and students. It supports workflows from contract review and privacy assessments to litigation and study, with attorney review required for legal outputs.
open_deep_research: Build Configurable Deep Research Agents
A Python research agent that searches the web, synthesizes findings, and produces reports using configurable language models and search tools. It suits developers who want to customize or deploy a research workflow with LangGraph.
firecrawl: Search, Scrape, and Crawl the Web for AI
Firecrawl turns web pages and sites into content that AI applications can use, including Markdown and structured data. It suits developers building agents or data pipelines that need web search, extraction, and crawling through an API or self-hosted deployment.
cherry-studio: Chat Across AI Models in a Desktop App
Cherry Studio is a cross-platform desktop client for chatting with cloud and local language models in one place. It suits people who want to compare models, use ready-made assistants, and work with documents without setting up a separate client for each provider.
mongodb-mcp-server: Connect AI Agents to MongoDB
A TypeScript Model Context Protocol server that lets AI agents query MongoDB databases and manage MongoDB Atlas resources. It suits developers who want database access through MCP, with read-only mode available for safer exploration.
context7: Give AI Coding Tools Current Library Docs
Context7 retrieves version-aware library documentation and code examples for AI coding assistants, helping reduce outdated guidance and invented APIs. It offers both a CLI-and-skills setup and an MCP server integration.