awesome-remote s
Remote MCP Servers
claude mcp add --transport stdio jaw9c-awesome-remote-mcp-servers npx -y jaw9c-awesome-remote-mcp-servers \ --env MCP_API_KEY="your-api-key" \ --env MCP_BASE_URL="https://example.com"
How to use
Awesome Remote MCP Servers is a curated directory of externally hosted MCP services that you can connect to from your AI applications. This repository itself lists official and well-maintained remote MCP endpoints, each exposed over the internet and accessible via standard MCP client integrations. To use these, you typically point your MCP client or LLM integration at the server URL provided in the list, and authenticate using the supported method (OAuth 2.1, API Key, or Open access). The documentation here highlights authentication options and quality criteria to help you choose a reliable, production-ready remote MCP server for your workflows.
When integrating, you can use either an MCP-ready client (such as Claude, ChatGPT, Cursor, or equivalent) or an API request flow to your chosen provider. For each server in the list, you’ll encounter a URL and an indicated authentication scheme. Depending on the server, you may need to register an app, obtain an API key, or complete an OAuth flow. After setup, your LLM can issue MCP calls to fetch data, run tools, or access external capabilities defined by the remote server.
How to install
Prerequisites:
- Access to a terminal with Git installed
- Basic knowledge of your chosen MCP client or integration method
Option A: Explore directly (no local installation required)
- Browse the Awesome Remote MCP Servers README to pick a server and copy its URL and authentication method.
- Use your MCP client to connect to the server URL using the recommended authentication flow (OAuth 2.1 or API Key).
Option B: Local integration skeleton (for developers)
- Create a project directory and initialize a minimal environment for your language of choice.
- If using Node.js, install an MCP client library (if applicable) and configure it to point at the chosen remote server URL with the required auth key.
- Implement a simple test call to verify you can reach the remote MCP server and perform a basic MCP operation (e.g., a read or tool invocation).
Example (pseudo-steps):
- git clone https://github.com/your-org/awesome-remote-mcp-servers.git
- cd awesome-remote-mcp-servers
- Configure environment variables as described in the server’s entry (MCP_API_KEY, etc.)
- Run your MCP client against the server URL and confirm a successful response.
Additional notes
Tips and common considerations:
- Always verify the server URL in the README and ensure the domain is trusted before connecting.
- Check the authentication method supported by the remote server (OAuth 2.1 vs API Key) and set up credentials accordingly.
- Use environment variables to securely store API keys or tokens rather than hard-coding them.
- If you encounter connectivity issues, verify network access to the server URL, and ensure any required OAuth redirect URIs or API key scopes are correctly configured.
- This repository focuses on listing servers; for detailed capabilities of a specific MCP server (endpoints, tools exposed, and response formats), refer to the individual server’s documentation linked in the table.