Where to Find a Curated List of MCP Servers

The definitive curated list of MCP servers is maintained in the README.md of the punkpeye/awesome-mcp-servers repository, with a synchronized web directory available at glama.ai/mcp/servers.

The punkpeye/awesome-mcp-servers repository serves as the central hub for discovering production-ready and experimental Model Context Protocol (MCP) server implementations. This curated list of MCP servers categorizes implementations by purpose, programming language, and deployment model, providing developers with a comprehensive resource for extending AI assistant capabilities.

Primary Sources for the Curated List of MCP Servers

The project maintains two synchronized sources for accessing the server catalog.

GitHub Repository (README.md)

The master curated list of MCP servers lives in the repository's top-level documentation. In /README.md (lines 16-25), you will find the authoritative catalog organized into thematic sections with visual indicators for quick scanning.

Access the raw list directly at:

https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md

Web-Based Directory (Glama.ai)

For users preferring a browseable interface, the repository points to a continuously synchronized web directory that mirrors the GitHub data. As noted in /README.md (lines 73-77), this alternative presentation maintains parity with the curated source while offering search and filter capabilities.

Visit the web directory at:

https://glama.ai/mcp/servers

Understanding the Curation Metadata

The curated list of MCP servers employs a standardized iconography system to convey critical implementation details at a glance:

• 🎖️ Official implementations – Servers maintained by the original service providers
• Language indicators – 🐍 Python, 📇 TypeScript/JavaScript, and others marking the primary runtime
• Deployment scope – ☁️ Cloud-hosted versus 🏠 Local-only installations
• Platform compatibility – 🍎 macOS, 🪟 Windows, 🐧 Linux support indicators

These visual markers appear consistently across both the GitHub README and the Glama.ai web interface, enabling rapid identification of suitable servers for specific use cases.

Programmatic Access to the MCP Server Catalog

Developers can consume the curated list of MCP servers through several programmatic approaches.

Fetching the Raw README with Python

Extract the server entries directly from the source documentation:

import requests
import re

url = "https://raw.githubusercontent.com/punkpeye/awesome-mcp-servers/main/README.md"
resp = requests.get(url)
readme = resp.text

# Extract all server entries (lines starting with dash)

servers = re.findall(r"^-\s+\[.*?\]\(.*?\).*", readme, re.MULTILINE)
for s in servers[:5]:  # Display first five entries

    print(s)

Querying the Web Directory via Bash

Locate specific implementations using standard command-line tools:

curl -s https://glama.ai/mcp/servers | grep -i "mindsdb"

Consuming JSON Data with JavaScript

Access structured metadata when available from the web directory:

fetch("https://glama.ai/mcp/servers/manifest.json")
  .then(r => r.json())
  .then(data => console.log(data.servers.filter(s => s.language === "🐍")))
  .catch(console.error);

Contributing to the Curated List

The repository welcomes community submissions to maintain the curated list of MCP servers. The /CONTRIBUTING.md file defines the submission guidelines, quality standards, and formatting requirements for adding new server implementations. All additions undergo manual review to ensure the list remains accurate and valuable.

For international users, localized versions such as /README-zh.md provide translated access to the same curated catalog, following identical organizational structures.

Summary

• The primary curated list of MCP servers is maintained in /README.md within the punkpeye/awesome-mcp-servers repository (lines 16-25).
• A synchronized web mirror exists at glama.ai/mcp/servers for browseable access (referenced in lines 73-77).
• The list uses standardized icons to indicate official status, programming language, deployment scope, and OS compatibility.
• Programmatic access is available via raw GitHub content, web scraping, or JSON APIs.
• Community contributions are governed by /CONTRIBUTING.md with support for multiple language variants.

Frequently Asked Questions

Where is the official curated list of MCP servers maintained?

The official curated list of MCP servers is maintained in the README.md file of the punkpeye/awesome-mcp-servers GitHub repository. This file contains the master catalog of server implementations, organized by category and annotated with compatibility indicators.

How can I access the MCP server list without using GitHub?

You can access a synchronized version of the curated list at glama.ai/mcp/servers, which mirrors the GitHub repository content. This web directory provides the same categorization and metadata as the GitHub source, accessible through any standard web browser.

What do the different icons in the server list represent?

The curated list uses specific emoji indicators: 🎖️ denotes official implementations maintained by service providers, 🐍 and 📇 represent Python and TypeScript/JavaScript respectively, ☁️ and 🏠 indicate cloud versus local deployment, while 🍎, 🪟, and 🐧 show macOS, Windows, and Linux compatibility.

How do I add my MCP server to the curated list?

Submit your server by following the guidelines in /CONTRIBUTING.md within the punkpeye/awesome-mcp-servers repository. The project requires specific formatting, working documentation links, and adherence to quality standards before inclusion in the curated list.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}