A customer messages support: checkout feels slow today. You open Scout, find the endpoint, open a trace, spot the query, copy it into Claude, and ask what’s wrong with it. Each step is quick, but together they add up to a lot of tab switching to answer one question.
Scout Monitoring is now available in Claude’s connector directory. Connect it once and Claude can look at your production data directly. Ask “why is checkout slow today?” and Claude pulls the endpoint’s response times from Scout, opens a slow trace, and shows you where the time went, down to the SQL.
Connecting Takes About a Minute
The connector works in Claude on the web and in the desktop app. Open the Scout Monitoring listing, click Connect, and sign in with your Scout account. There are no API keys to copy and nothing to install.
Sign-in uses OAuth, so your credentials are never shared with Claude, and Claude can only see the organizations and apps your Scout account can access. There’s no extra charge for the connector. It uses the API access that comes with your existing Scout account.
What You Can Ask
The connector gives Claude 17 read-only tools covering the data Scout already collects for Ruby, Python, PHP, Elixir, and Node.js apps. In practice, that means you can ask questions like these and get answers grounded in your actual production data:
- “Which endpoints got slower this week, and why?”
- “Show me the N+1 queries Scout has found in our Rails app.”
- “Which errors are happening most often right now?”
- “Scout flagged an anomaly on the search endpoint. What happened?”
- “Which background jobs have the longest queue times?”
The answers come from Scout’s own analysis, which matters more than it might seem. Raw metrics tell an agent that an endpoint is slow. Scout’s insights for N+1 queries, slow queries, and memory bloat tell it why, so Claude can go from symptom to cause without guessing. From there it can explain the problem and suggest a fix, and if you work in Claude Code, apply that fix in your codebase.
Read-Only by Design
Every tool in the connector reads data. None of them can change your Scout account, resolve errors, edit alerts, or modify settings.
That was a deliberate choice. An AI assistant that can browse your production data is useful. One that can also change your monitoring configuration is a different kind of risk, and most teams would rather not think about it. With Scout, the worst an assistant can do is read something you already have access to.
Already Using Scout MCP Elsewhere?
Nothing changes. The directory listing uses the same hosted server at https://scoutapm.com/mcp that works with Claude Code, Cursor, VS Code with Copilot, and other MCP clients. In Claude Code, it’s still one command:
claude mcp add --transport http scout https://scoutapm.com/mcp
The Scout MCP page has setup steps for every client, plus a self-hosted option if you’d rather run the server yourself.
Try It on Your Next Slow Endpoint
Next time someone says something feels slow, start by asking Claude instead of opening a dashboard. If the answer doesn’t match what you’d have found yourself, we want to hear about it at support@scoutapm.com.
New to Scout? Start monitoring for free and connect Claude in the same afternoon.
For application monitoring with errors, logs, and traces, Scout Monitoring provides the fastest insights without the bloat.