MercurALPHA
Sign inCreate a free account

Debug webhooks with Cursor MCP

When a webhook fails, the payload that actually hit Mercur is already in request history. With remote MCP, an MCP client (Cursor, Claude Desktop, and others) can list tunnels and open those requests without you copy-pasting from the console.

Prerequisites

  • A Mercur workspace with API keys enabled
  • At least one HTTP local tunnel with logged requests
  • An MCP client that supports remote / HTTP servers

Create a workspace API key

In the console open Settings → Workspace → API keys. Any member can create a key; copy the secret and MCP config snippet once.

Open Workspace Settings

Add the remote MCP server in your client

Paste the snippet into your MCP client as a remote / HTTP server so it calls console …/api/mcp with Authorization: Bearer mcr_… The snippet shape is:

{
  "mcpServers": {
    "mercur": {
      "url": "https://console.mercur.sh/api/mcp",
      "headers": {
        "Authorization": "Bearer mcr_…"
      }
    }
  }
}

On localhost the URL uses your console origin instead.

List servers and inspect requests

Ask the agent to call list_servers, then list_requests / get_request for a failed webhook. Summaries omit bodies; detail includes headers and bodies. Tools are scoped to the workspace that owns the key — they cannot read another workspace’s tunnels.

Related