Connecting your agent (MCP)

AgentLeverage exposes your organization's tools through a remote MCP server (Model Context Protocol). Once connected, an AI agent — Claude Code, Cursor, Claude Desktop, and others — can call your org's tools directly in your organization.

An API token is a separate principal. Jobs it creates appear in Jobs (badged as the token / agent), but the token cannot read jobs you started in the app.

Find your connection details

Your endpoint and copy-paste config are specific to your organization and are always available in the app:

Agents — in the sidebar

That page shows your exact endpoint URL and a ready-to-copy config for each supported client. The production endpoint is:

https://www.agentleverage.co/api/mcp/mcp

Authentication

Connections authenticate with a bearer token sent as an Authorization header. Each client below shows where the token goes.

Claude Code

Claude Code connects to remote MCP servers natively. Add the server to your .mcp.json (project scope) or ~/.claude.json (user scope), then restart Claude Code:

{
  "mcpServers": {
    "agentleverage": {
      "type": "http",
      "url": "https://www.agentleverage.co/api/mcp/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_AGENTLEVERAGE_TOKEN>"
      }
    }
  }
}

The "type": "http" field is required for remote servers.

Cursor

Add the server to ~/.cursor/mcp.json (global) or .cursor/mcp.json (per-project). Set an AGENTLEVERAGE_TOKEN environment variable, then restart Cursor:

{
  "mcpServers": {
    "agentleverage": {
      "url": "https://www.agentleverage.co/api/mcp/mcp",
      "headers": {
        "Authorization": "Bearer ${env:AGENTLEVERAGE_TOKEN}"
      }
    }
  }
}

Cursor auto-detects the transport from the URL — no type field is needed.

Claude Desktop

Claude Desktop doesn't connect to remote servers directly, so it uses the mcp-remote bridge. Add this to claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/; Windows: %APPDATA%\Claude\), then fully quit and relaunch Claude Desktop:

{
  "mcpServers": {
    "agentleverage": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://www.agentleverage.co/api/mcp/mcp",
        "--transport",
        "http-only",
        "--header",
        "Authorization:${AUTH_HEADER}"
      ],
      "env": {
        "AUTH_HEADER": "Bearer <YOUR_AGENTLEVERAGE_TOKEN>"
      }
    }
  }
}

ChatGPT

ChatGPT is not yet supported. ChatGPT's Developer-mode connectors require OAuth 2.1 and can't accept a static bearer token, so there's no way to paste your AgentLeverage credentials. Support is planned alongside upcoming OAuth work.

What your agent can do

Once connected, your agent sees your organization's tools in its tool list. It calls them as the API token's separate principal. Jobs it creates appear in Jobs (badged as the token / agent); the token still cannot read jobs you started in the app. Your agent can also check your org's credit balance (get_credit_balance) and list, read, or cancel its own jobs (list_jobs, get_job, cancel_job). If your organization proxies additional upstream MCP servers, those tools appear too, namespaced per server.

Admins can review and manage the full tool catalog under Settings → MCP servers.

Troubleshooting