Connect a client

Connect other MCP clients to OClever

Set up OClever in Windsurf, Gemini CLI, Microsoft Copilot Studio or any client that speaks MCP over HTTP.

Any MCP client

OClever is a remote MCP server using the Streamable HTTP transport. Any client that supports remote servers can connect with:

  • URL: https://mcp.oclever.com/mcp
  • Authentication: OAuth (discovered automatically from the server), or a bearer token header.
mcp.json
{
  "mcpServers": {
    "oclever": {
      "type": "http",
      "url": "https://mcp.oclever.com/mcp",
      "headers": {
        "Authorization": "Bearer oc_pat_your_token_here"
      }
    }
  }
}

Field names differ between clients (url, serverUrl, httpUrl); check your client's docs for remote servers. Omit headers if the client handles OAuth sign-in.

Windsurf

Open Windsurf Settings → Cascade → MCP servers and choose View raw config, or edit the file directly:

~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "oclever": {
      "serverUrl": "https://mcp.oclever.com/mcp",
      "headers": {
        "Authorization": "Bearer ${env:OCLEVER_TOKEN}"
      }
    }
  }
}

Refresh the MCP list in Cascade after saving.

Gemini CLI

shell
gemini mcp add --transport http oclever https://mcp.oclever.com/mcp

Then run /mcp auth oclever in the CLI to sign in. To use a token instead, add it to your settings file:

~/.gemini/settings.json
{
  "mcpServers": {
    "oclever": {
      "httpUrl": "https://mcp.oclever.com/mcp",
      "headers": {
        "Authorization": "Bearer $OCLEVER_TOKEN"
      }
    }
  }
}

Microsoft Copilot Studio

  1. 1
    Open your agent and go to Tools → Add a tool → New tool → Model Context Protocol.
  2. 2
    Enter the name OClever, a description such as “AI visibility reports from OClever”, and the server URL.
  3. 3
    For authentication choose OAuth 2.0 with dynamic discovery, or API key as a header named Authorization with the value Bearer oc_pat_….
  4. 4
    Create the connection, sign in, and add the tool to your agent.

Clients that only support local servers

Some clients can only launch local (stdio) servers. Bridge them to OClever with the open-source mcp-remote package (needs Node.js):

mcp.json
{
  "mcpServers": {
    "oclever": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.oclever.com/mcp",
        "--header",
        "Authorization:Bearer ${OCLEVER_TOKEN}"
      ],
      "env": {
        "OCLEVER_TOKEN": "oc_pat_your_token_here"
      }
    }
  }
}

Without the --header arguments, mcp-remote runs the OAuth sign-in in your browser instead.

Testing with MCP Inspector

To see exactly what the server returns, run the official inspector, choose Streamable HTTP, paste the URL and connect:

shell
npx @modelcontextprotocol/inspector

Create tokens for scripts and test tools at Settings → AI assistants.

Last updated 5 October 2026. Questions or something unclear? Email [email protected].