Skip to main content
MCP is the easiest way to use OpenRush. After you connect OpenRush, your AI agent can call OpenRush tools for you. You can ask normal questions about competitors, keywords, search results, and website issues.
MCP is the bridge between your agent and OpenRush data.
Use Setup in the Dashboard for the current MCP URL and client-specific instructions.

​
What MCP does

MCP lets your AI tool ask OpenRush for data. For example, you can ask:
Use OpenRush to analyze example.com. Find competitors, keyword gaps, and the next pages I should inspect.
Your agent can then call OpenRush tools behind the scenes and explain the results in plain language.

​
Claude.ai

  1. Open Customize.
  2. Open Connectors.
  3. Click Add custom connector.
  4. Paste the OpenRush MCP URL from Setup.
  5. Click Connect and approve access.

​
Other AI clients

If your AI client supports remote MCP servers or custom connectors, use the MCP URL from Setup. This includes tools that accept a JSON MCP config or an HTTP MCP server URL. The exact button names vary by product, so use your client’s connector or MCP settings.

​
Claude Code

Use the command shown in Setup. It follows this shape: 
claude mcp add --transport http openrush https://api.openrush.io/mcp
claude mcp login openrush

​
Codex

Add the OpenRush MCP server to your Codex config: 
[mcp_servers.openrush]
url = "https://api.openrush.io/mcp"
Then run: 
codex mcp login openrush

​
Direct API access

Developers can skip MCP and call the API directly. Use API keys from API access: 
curl -X POST "https://api.openrush.io/v1/tools/inspect_domain" \
  -H "Authorization: Bearer $OPENRUSH_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain":"stripe.com"}'
MCP and API key calls share the same account credits. Review recent requests in Activity.