The official Sentry MCP server pulls live error data directly into your AI agent. Instead of screenshotting stack traces and pasting them into chat, point Claude at a Sentry issue — it gets full context, correlates with releases, and suggests fixes.
Features:
- Query issues by project, environment, and time range
- Get full stack traces with local variables
- Access performance transaction data and bottlenecks
- Correlate errors with recent deployments and releases
- Search events by user, URL, tag, or error type
- Manage issue status (resolve, ignore, merge)
- Access release health metrics
- Set up alert rules and notification policies
The typical workflow:
1. Claude sees an error in your code
2. Claude queries Sentry for matching recent issues
3. Pulls the full stack trace, breadcrumbs, and affected users
4. Suggests a fix with full context — no manual copy-paste
Built and maintained by Sentry. OAuth-based — no stored API keys.
sentry
errors
monitoring
debugging
observability
performance