Scout Mcp Local
Let your coding agent query Scout APM metrics and monitoring data while you debug production issues from the terminal.
Overview
com.scoutapm/scout-mcp-local is an Operate-phase MCP server that lets AI agents interact with Scout Monitoring data via stdio after you configure SCOUT_API_KEY.
What is this MCP server?
- Stdio MCP via PyPI package scout-mcp-local or Docker image scoutapp/scout-mcp-local
- Authenticated access with required SCOUT_API_KEY environment variable
- Bridges Scout Monitoring product data into Claude Code, Cursor, and other MCP clients
- Local stdio transport—no remote HTTP endpoint in server manifest
- Version 2025.11.4 aligned with official server.schema.json 2025-10-17
- Server version 2025.11.4
- Two distribution paths: PyPI scout-mcp-local and Docker scoutapp/scout-mcp-local:latest
- One required secret: SCOUT_API_KEY
What problem does it solve?
You are triaging production slowness in your agent session but still have to leave the chat to dig through Scout’s UI for the same metrics.
Who is it for?
Indie SaaS builders already on Scout APM who want MCP-native access to monitoring data from Claude Code or Cursor.
Skip if: Teams without Scout Monitoring, or anyone who only needs local log tailing without an APM product.
What do I get? / Deliverables
Your agent can pull Scout Monitoring context on demand so debugging and monitoring workflows stay in one place.
- Registered stdio MCP server wired to Scout Monitoring
- Agent-accessible Scout monitoring interactions per server capabilities
- Repeatable local or containerized MCP deployment for your team
Recommended MCP Servers
Journey fit
Production observability sits in Operate—after you ship, you need app performance and error signals without leaving the agent workflow. Monitoring is the canonical shelf for APM-backed MCP tools that expose traces, throughput, and health data to agents.
How it compares
Scout APM data bridge via MCP, not a standalone error-tracking skill or generic infra provisioner.
Common Questions / FAQ
Who is com.scoutapm/scout-mcp-local for?
Solo and small-team builders who use Scout Monitoring and want their AI coding agent to read or reason over that monitoring data without manual copy-paste from the dashboard.
When should I use com.scoutapm/scout-mcp-local?
Use it during Operate and incident debugging when you already rely on Scout and want stdio MCP access while fixing performance or availability issues in production.
How do I add com.scoutapm/scout-mcp-local to my agent?
Install scout-mcp-local from PyPI or run the published Docker image, set the required SCOUT_API_KEY secret in the environment, and register the server as an stdio MCP entry in Claude Code, Cursor, or another compatible client.