Scout’s MCP is the Al-native solution that exposes performance issues and slow endpoints to your coding assistants and makes monitoring in plain language possible.
Works with all AI assistants supporting MCP
Bugs after deploy that are hard to explain or reproduce
APM + errors + logs = separate, heavy, expensive
Context-switching kills momentum, and UX suffers
Skip dashboards and get clear performance monitoring answers on demand
With insights from Scout MCP, your agents learn to suggest contextual fixes for any detected issues.
A local MCP server allowing Scout users to explore their application monitoring metrics, errors, and logs within their coding agent.
Surface AI-specific bugs before they hit production
Errors, Logs, and Traces – right where you need them
Focus on informed delivery, no context-switching needed
Anything that can work with Model Context Protocol running locally. Currently, this includes Claude Desktop, Claude Code, Cursor, VS Code / Copilot, Windsurf, Cline/Roo, Amp, “That Thing That Was Just Released This Morning”… more than we can keep up with, which is why MCP exists, right?
Yes.
There are no associated costs related to Scout. The local server relies on the API access included in your Scout Monitoring account. It is only as good as the data you send, so being in the correct tier to send all of your transaction, error and log data will benefit you. API rate limits may apply.Of course, calls to the Scout MCP will result in tokens being processed by your AI Assistant of choice, so there is an indirect impact.
The Scout MCP server has access to insight, error, metric, and trace data that the Scout Agent collects from your application. It is currently read-only access so the MCP server cannot change anything about your Scout Monitoring account. But it can help you troubleshoot N+1’s, fix errors and improve app performance by putting runtime data within reach of your favorite AI assistant or coding companion. Log data will be available soon.
