Skip to main content

Connecting AI tools

Every client needs the endpoint URL, such as https://docs.example.com/mcp, and a name for the server. Replace my-docs and the URL below with yours. To connect to these docs, use https://docusaurus-plugin-mcp-server.vercel.app/mcp.

The install button generates all of these for your readers, with your URL and name filled in.

Claude Code​

claude mcp add --transport http my-docs https://docs.example.com/mcp

Add --scope user to make it available in every project, or --scope project to write it to the project's .mcp.json and share it with your team.

Cursor​

Add to ~/.cursor/mcp.json, or .cursor/mcp.json in a project:

{
  "mcpServers": {
    "my-docs": {
      "url": "https://docs.example.com/mcp"
    }
  }
}

VS Code​

code --add-mcp '{"name":"my-docs","type":"http","url":"https://docs.example.com/mcp"}'

Or add it to .vscode/mcp.json in a workspace. VS Code uses servers, not mcpServers:

.vscode/mcp.json
{
  "servers": {
    "my-docs": {
      "type": "http",
      "url": "https://docs.example.com/mcp"
    }
  }
}

Codex​

codex mcp add --url https://docs.example.com/mcp my-docs

Or in ~/.codex/config.toml:

~/.codex/config.toml
[mcp_servers.my-docs]
url = "https://docs.example.com/mcp"

Gemini CLI​

gemini mcp add --transport http my-docs https://docs.example.com/mcp

Windsurf​

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "my-docs": {
      "serverUrl": "https://docs.example.com/mcp"
    }
  }
}

Claude Desktop and claude.ai​

Add the URL as a custom connector in Settings → Connectors. For a config file setup without connectors, Claude Desktop can reach a remote server through mcp-remote:

claude_desktop_config.json
{
  "mcpServers": {
    "my-docs": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://docs.example.com/mcp"]
    }
  }
}

Other clients​

Any client that supports the Streamable HTTP transport can connect with just the URL. The endpoint needs no authentication or session setup.

What to ask​

Once connected, ask questions your docs answer. The agent calls docs_search, then docs_fetch on the best matches, and cites the pages it used. Clients that support the skills extension also load the docs-research skill, which walks the agent through that workflow.