MCP.so is a directory for discovering Model Context Protocol servers, clients, CLI tools, skills, and agents that connect AI apps like Claude, ChatGPT, and Cursor to external tools and data.
About
MCP.so is a directory website for the Model Context Protocol (MCP), the open standard that lets AI applications such as Claude, ChatGPT, and Cursor connect to external tools, data sources, and automated workflows. The site addresses a real discovery problem: as the number of MCP servers and clients has grown, developers need a organized place to find production-ready options rather than searching scattered repositories.
The site organizes its listings into sections for servers, remote servers, clients, CLI tools, loops (reusable agent workflows), skills, and agents, plus categories, tags, a submission feed, and rankings. Visitors can browse curated categories such as Developer Tools, AI & Agents, Cloud & Infrastructure, Memory & Knowledge, and Media & Design. The site also publishes guides on MCP setup, authentication, security, and tooling through its blog. Developers can submit their own open-source MCP servers for inclusion, and the directory highlights hand-picked featured entries alongside trending community picks.
Key Features
Pricing & Fees
- Pricing: Not publicly listed on the pages we reviewed; check the official site for current plans. A pricing page exists at mcp.so/pricing, and the site carries advertising, suggesting free browsing with paid promotional placement for listed products.
- Platform coverage: Listings span tools for AI apps including Claude, ChatGPT, and Cursor, as well as general MCP clients.
- Content types: Beyond servers and clients, the directory covers CLI tools, loops, skills, and AI agents, giving a broad view of the MCP ecosystem.
- Curation: Featured servers are described as hand-picked and production-ready, and a trending section shows what the community is installing each week.
Who It's For
MCP.so is aimed primarily at software developers and engineers building or configuring AI agents that use the Model Context Protocol. Typical users include developers wiring an MCP server into Claude Code, Cursor, or ChatGPT; DevOps engineers connecting agents to cloud infrastructure like Cloudflare; and builders looking for memory, database, or web-data servers to extend their assistants. It is also useful for maintainers of open-source MCP projects who want exposure through the directory's submission and feed system, and for anyone comparing MCP tooling before committing to an integration.
FAQ
What is MCP.so?MCP.so is a directory for discovering MCP servers, clients, CLI tools, skills, and AI agents that connect AI applications like Claude, ChatGPT, and Cursor to external tools and data.
Is MCP.so free to use?Browsing the directory appears free; pricing plans are not publicly detailed on the pages reviewed, and the site offers a separate pricing page and advertising options for those who want to promote listings.
Which AI apps and platforms does MCP support?The Model Context Protocol is used by AI apps including Claude, ChatGPT, and Cursor, and the directory lists many clients, IDEs, and agent frameworks that speak MCP.
Can developers submit their own MCP servers to MCP.so?Yes, the site has a submission page where developers can submit open-source MCP servers for inclusion in the directory and its feed.
Does MCP.so offer setup guides?Yes, its blog includes step-by-step guides for adding MCP servers to Claude Code, ChatGPT, and OpenAI Codex, along with explainers on MCP authentication, security, and tool discovery.
Related Tutorials
View all →
Related Tools
View all →
The Agent Skills Directory Explore AI agent skills from Microsoft and GitHub. Discover, compare, and integrate capabilities to enhance your AI workflows. 
Orca Orca is a free, open-source Agent Development Environment that runs CLI coding agents like Claude Code, Codex, and Gemini in parallel across isolated git worktrees. 
Cursor Cursor is an AI coding agent and development environment that turns ideas into working code, assisting software engineers and teams across desktop, terminal, and cloud workflows. 
Kilo Kilo Code is an open source AI coding agent for VS Code, JetBrains IDEs, the CLI, and the cloud, offering 500+ models with zero markup and local model support. 
HumanLayer HumanLayer is a multiplayer coding agent IDE and cloud workspace where engineering teams run AI agent sessions, review plans, and collaborate on code diffs in one place.