Skip to main content
Connect large language models (LLMs) to Chronosphere Observability Platform using the Model Context Protocol (MCP) server. After connecting, your LLM can:
The Chronosphere MCP server is read-only. It doesn’t support mutation actions such as updating dashboards, writing metrics and logs, or deleting monitors.
Use AI-enabled developer tools such as Claude Code, Cursor, Codex, and Gemini to submit queries. Example use cases:
  • Find error logs with stack traces and review code for root cause analysis.
  • Fetch a monitor’s configuration and explain what’s triggering it.
  • Write PromQL queries targeting a specific service’s metrics.
  • List active shaping rules, such as drop rules, to determine why data is or isn’t being dropped.

Get started

To get started with the Chronosphere MCP server, configure authentication.