Sentry Issue Investigation

SkillMonitoring & ops

This skill walks your AI through investigating a Sentry error report before it proposes any code change. It draws on the issue's events, stack traces, tags, environments, releases, and timeline as evidence. The result is a suggested fix based on what the issue data actually shows.

Available today. Use it from your connected AI after setup.

After adding the skill, give your AI a Sentry issue and ask it to investigate the error before proposing a fix.

Then ask your AI: use the Sentry Issue Investigation skill

What your AI can do with it

  • Investigate a Sentry issue step by step before suggesting a code fix
  • Review event details and stack traces tied to an error
  • Check tags, environments, and releases connected to the issue
  • Trace the issue's timeline to see how the error developed
  • Propose code changes grounded in the collected evidence

What this skill tells your AI

The instructions your AI receives, as published by devin-axis/ipollowork in examples/plugin-packages/sentry/skills/sentry-issue-investigation/SKILL.md and read by ahel’s review.

  1. Resolve the organization, project, environment, issue, and time window.
  2. Read representative recent events rather than assuming the issue summary is complete.
  3. Compare stack frames, exception values, tags, releases, affected users, and recurrence patterns.
  4. Separate the observed failure, likely trigger, suspected code path, and remaining uncertainty.
  5. When implementation is requested, inspect the matching code and make the smallest evidence-backed fix.

Do not expose user data from event payloads. Do not claim root cause from a title or one stack trace alone.

Signals

GitHub stars
6k
Forks
1k
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
sentry-issue-investigation
Source
github.com/devin-axis/ipollowork