FlareLog logoFlareLog
// AI DEBUGGING

Your AI can't fix what it can't see

You built an app with Cursor or Claude. It worked locally. Then production broke. You paste the error into your AI, but it only sees what you pasted. Not your logs. Not the request that caused it. Not the 50 other errors that happened in the same minute.

The old way (broken)

1. App crashes in production
2. You see "500 Internal Server Error"
3. You open Cloudflare dashboard — "Error 1101, no logs"
4. You copy-paste the error into ChatGPT
5. ChatGPT guesses — "maybe it's a null reference?"
6. You change random code based on a guess
7. Deploy. Still broken. Repeat.

The FlareLog way (MCP)

1. App crashes in production
2. FlareLog's Tail Worker captures the full stack trace
3. You open Cursor or Claude Desktop
4. Ask: "What broke in production in the last hour?"
5. Cursor queries FlareLog via MCP — sees the actual error, stack trace, request URL
6. Cursor opens the file, highlights the bug, suggests a fix
7. Apply. Deploy. Fixed. Done.

Set up MCP in 2 minutes

For Cursor

Create .cursor/mcp.json in your project:

{
  "mcpServers": {
    "flarelog": {
      "url": "https://mcp.flarelog.dev",
      "headers": {
        "Authorization": "Bearer fl_your_api_key_here"
      }
    }
  }
}

For Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "flarelog": {
      "url": "https://mcp.flarelog.dev",
      "headers": {
        "Authorization": "Bearer fl_your_api_key_here"
      }
    }
  }
}

What you can ask your AI

  • ›"What errors happened in production in the last hour?"
  • ›"Show me the stack trace for the crash that happened at 3:42 AM"
  • ›"Which API endpoint is throwing the most 500s?"
  • ›"Find the request that caused this traceId and suggest a fix"
  • ›"Are there any patterns in the errors from the last deploy?"

FAQ

Can I use an MCP server for debugging production errors with AI?

Yes — that's exactly what FlareLog's MCP server does. Connect it to Cursor, Claude Desktop, Claude Code, Windsurf, or Lovable Agent, then ask in plain English: 'What broke in production in the last hour?' Your AI queries your live logs, reads stack traces with full request context, and suggests fixes. No copy-pasting error messages.

Which AI tools work with the FlareLog MCP server?

Any MCP-compatible client: Cursor (.cursor/mcp.json), Claude Desktop and Claude Code, Windsurf, Lovable Agent, plus any tool that implements the MCP client standard. Setup is one JSON block with your FlareLog API key — about 2 minutes. ChatGPT does not support MCP natively, so you'd use the FlareLog dashboard directly.

What is an MCP server?

MCP (Model Context Protocol) is a standard that lets AI tools like Cursor and Claude Desktop read external data sources. FlareLog's MCP server exposes your production logs, errors, and traces to your AI, so instead of copy-pasting error messages, your AI queries your dashboard directly.

Is my production data safe with the MCP server?

Yes. The MCP server authenticates with your FlareLog API key, scoped per project. Your AI can only read logs you've already ingested. It can't modify or delete anything. You can revoke access at any time by rotating your API key.

Can the AI actually fix bugs, or just find them?

Both. The AI reads the error, sees the stack trace, correlates it with recent deploys, and suggests a fix. In Cursor, it can open the file and apply the fix directly. In Claude Desktop, it gives you the code to paste. The quality depends on the model — Claude Sonnet and GPT-4 are excellent at this.

Let your AI see production

Free 10,000 logs/mo. MCP server included. No credit card.

Start free