Sentry

Investigate errors and performance issues with Seer.

Observability

Investigate and triage production errors with Seer-powered root-cause analysis, without leaving your editor.

The Sentry MCP server connects AI coding assistants directly to your Sentry organization, letting them pull live error, issue, trace, and release data instead of copy-pasting stack traces between Sentry and your editor. It integrates Seer — Sentry's AI agent — to perform automated root-cause analysis across stack traces, traces, logs, and your codebase, and propose a fix. Available as a hosted remote server or run locally.

Transport: http
Authentication: OAuth
Endpoint: https://mcp.sentry.dev/mcp

Authenticates with your Sentry organization via OAuth on first connect.

What you can do

Investigate an issue's root cause with Seer
List and search issues and error events
Search events and traces in natural language
Inspect stack traces, activity, and details
Browse projects, teams, and releases
Update issue status and project settings
Pull replay, profile, and trace details

Try asking

“What are the most frequent unresolved issues in production this week?”

“Use Seer to find the root cause of issue PROJECT-123 and suggest a fix.”

“Show me the stack trace and recent events for the latest checkout error.”

Tools it exposes

A sample of the tools available once connected.

search_issuesget_issue_detailssearch_eventsanalyze_issue_with_seerget_trace_detailsfind_releases

Set it up

One click adds Sentry to Cursor. Cursor must be installed.

Add to Cursor

Official Sentry MCP docs