sentry-audit

name	sentry-audit
description	Fetches unresolved Sentry issues and maps stack traces to local codebase for ShizukuPlus.
license	Complete terms in LICENSE
metadata	{"author":"Google LLC","keywords":["sentry","audit","crashes","debugging"]}

Sentry Audit Skill

This skill minimizes token usage while extracting actionable crash data from the Sentry REST API and mapping it to the local codebase.

Execution Steps:

1. Fetch Unresolved Issues: Run a curl command against the Sentry API to fetch unresolved issues. Extract ONLY the id, shortId, title, and culprit.

   curl -s -H 'Authorization: Bearer <SENTRY_TOKEN>' "https://sentry.io/api/0/projects/af-developments/shizukuplus/issues/?query=is:unresolved" | jq -c '.[] | {id, shortId, title, culprit}'
   

   Note: Do NOT fetch the full JSON dump into context; it is massive and wastes tokens.

2. Fetch Stack Trace for Top Issue: Pick the highest priority issue ID and fetch its exact stack trace.

   curl -s -H 'Authorization: Bearer <SENTRY_TOKEN>' "https://sentry.io/api/0/issues/<ISSUE_ID>/events/latest/" | jq -r '.entries[] | select(.type=="exception") | .data.values[].stacktrace.frames[] | "\(.filename):\(.lineNo) - \(.function)"'
   

   Note: Using jq filters the payload down to just filename, line number, and function, drastically saving context space.

3. Map to Codebase: Use the grep_search tool to find the corresponding file in manager/src/main/ based on the .filename from the stack trace.

4. Repair: Analyze the bug and use the replace_file_content or multi_replace_file_content tool to patch the error.

5. Resolve via API: Once pushed to master, automatically resolve the issue on Sentry:

   curl -s -X PUT -H 'Authorization: Bearer <SENTRY_TOKEN>' -H "Content-Type: application/json" -d '{"status":"resolved"}' "https://sentry.io/api/0/issues/<ISSUE_ID>/"