Sentry
VerifiedRead-only issue, event, project, environment, and release evidence for bounded incident triage.
What you can do with Sentry
Turn a bounded issue set into a redacted incident brief with transparent ordering, release timing, evidence links, and explicit unknowns.
Useful workflows
- Prioritize unresolved issues by regression state, recency, affected users, and event count.
- Inspect one representative event after removing tokens, request bodies, email addresses, IP addresses, and user fields.
- Compare issue first-seen timestamps around an explicit release without claiming the release caused them.
Connection and permissions
Use a Sentry organization token with event:read and project:read; write and admin scopes are neither required nor used.
Pricing basis
Sentry does not publish a separate per-request API charge for these reads; ToolRouter charges the displayed skill price and failed paid calls do not charge.
Evidence limits
- One job is capped at ten projects, 100 issues, and a 30-day observation window.
- The tool does not download replays, attachments, source maps, or raw log archives.
- It never resolves, assigns, or edits an issue and never changes code, creates a ticket, or opens a pull request.
Tools (1)
Frequently Asked Questions
Can ToolRouter fix a Sentry issue automatically?
No. This tool is deliberately read-only: it organizes redacted evidence and next checks while leaving code and incident mutations to a separately authorized workflow.
How does Sentry event redaction work?
The inspection skill removes credential fields, authorization and cookie data, request bodies, email addresses, IP addresses, and user objects before returning event evidence.
Does an issue first seen after a release prove causation?
No. The comparison labels first-seen timing as correlation and keeps root cause unknown unless separate deployment and code evidence establishes it.