awesome-mcp-servers README

The project's README from github.com/mcpHQ/awesome-mcp-servers. As one markdown file: README.md.

Awesome MCP Servers

Servers GitHub stars Last commit Link check MCP License: MIT

A hand-curated, link-checked catalog of Model Context Protocol (MCP) servers, with an interactive map and a JSON API.

MCP Landscape: interactive, searchable map of every server in this catalog

Explore the live MCP Landscape →

MCP is an open protocol that lets AI applications connect to external tools and data through a standardized client-server interface. This list focuses on well-scoped, source-available servers that extend AI workflows with databases, developer tools, browsers, cloud services, and more.

Why this list

If this list saves you time, please ⭐ star the repo. It helps other people find it.

Catalog

Official and Reference Servers

Reference implementations, SDKs, and core protocol tooling from the MCP ecosystem.

Databases and Storage

Query, manage, and explore databases, vector stores, and data warehouses.

Developer Tools and Code Intelligence

Repositories, CI/CD, observability, and coding workflows for software teams.

Browsers, Search, and Web Automation

Browse the web, scrape content, search, and automate browser interactions.

Filesystems and Documents

Read and write files, convert documents, and connect to knowledge bases.

Cloud and Infrastructure

Manage cloud resources, deploy services, and operate infrastructure.

Communication and Productivity

Integrate chat, email, calendars, and team collaboration tools.

AI, Agents, and Memory

LLM bridges, agent orchestration, RAG, and persistent memory layers.

Data, Analytics, and BI

Analytics platforms, monitoring, and business intelligence tooling.

Court records, legal research, case search, and judicial workflow data.

Security and Identity

Secrets, vulnerability scanning, authentication, and security operations.

Finance, Commerce, and Business Apps

Payments, banking, CRM, e-commerce, and business system integrations.

Utilities and Examples

Helpful utilities, templates, and starter servers for learning MCP.

Use the Data

The full catalog is published with every landscape deploy:

File URL
Servers https://landscape.mcphq.org/api/servers.json
Categories https://landscape.mcphq.org/api/categories.json
LLM index https://landscape.mcphq.org/llms.txt
LLM full catalog https://landscape.mcphq.org/llms-full.txt
curl -s https://landscape.mcphq.org/api/servers.json | jq '.[] | select(.official) | .name'

The fields are described in data/servers.schema.json. Please link back to this repo if you build on the data.

Listed on mcpHQ Badge

If your server is in this catalog, you can add this badge to your README:

Listed on mcpHQ

[![Listed on mcpHQ](https://img.shields.io/badge/Listed%20on-mcpHQ-8A2BE2)](https://github.com/mcpHQ/awesome-mcp-servers)

Quality Criteria

An entry is accepted if the server is:

An entry is removed if its link stays broken, its repository is archived without a maintained successor, or it turns out to be misleading about what it does or who backs it.

Contribute

Found a great MCP server? Read CONTRIBUTING.md and open a PR with an entry in data/servers.json.

Then regenerate the README and LLM catalog files:

npm run generate
npm run validate

See landscape/README.md for how the landscape is built and how to preview or customize it locally.

License

This list is released under the MIT License.

awesome-mcp-servers (mcpHQ/awesome-mcp-servers)

Searching Database…

What is this?A hand-curated, link-checked catalog of Model Context Protocol (MCP) servers — tools that let AI applications connect to databases, developer tools, browsers, cloud services and more.
Is it just a README list, or is there something structured behind it?Structured data: every entry lives in data/servers.json with category, language, provider, tags and an official/community flag, and the README, llms.txt files and the landscape map are all generated from that file.
What is MCP in the first place?An open protocol that lets AI applications connect to external tools and data through a standardized client-server interface — any MCP-compatible client can talk to any MCP server.
How big is the catalog?About 178 servers across the 13 categories. The largest sections are AI, Agents, and Memory (24), Developer Tools and Code Intelligence (20), and Finance, Commerce, and Business Apps (21).
Is there a more visual way to explore it than scrolling the README?Yes — the repo links an interactive MCP Landscape map of the ecosystem, handy when you do not yet know which category you need. There is also a short LLM index built for agents.
Creator
mcpHQ
Size
3.6 MB
Stars
191
Updated
CommitOct 16
Age
4mo
Commits · 90 Days