Skip to content
← Tools / Skills
Skill · .md packageSentry/10 trustMetrics-checked
Sentry Skills logo

Sentry Skills

Official Agent Skills used by the Sentry team for development.

Visit official site →

About

Sentry Skills is the official agent-skills collection from the Sentry organization, maintained in the getsentry GitHub org under Apache-2.0 and following the Agent Skills open format (agentskills.io). It ships ~28 skills (code review, commit, PR writing, bug finding, Django security/perf reviews, GitHub Actions security review) plus 2 subagents (code-simplifier, senpai engineering mentor), including security-oriented ones like a skill-scanner that audits other skills for issues and a secret-serialization leak detector. The README points product setup/debugging skills (adding Sentry to a project, triaging production issues) to the separate getsentry/sentry-for-ai repo.

What it can do

  • 28 skills: code review, commit workflow, PR writing, bug finding, doc co-authoring
  • Security reviews: generic security review, GitHub Actions security review, Django access-control/IDOR review
  • Secret-serialization leak detection: finds secrets leaking through dataclass repr, asdict, model_dump, JSON.stringify, logs, tracing spans
  • skill-scanner: audits agent skills for security issues
  • Django performance review and typing-exclusion worker
  • 2 subagents: code-simplifier and senpai (senior-engineer mentor for new hires)
  • Skills follow the Agent Skills open specification with SPEC.md maintenance contracts

Setup

  1. In Claude Code, register the marketplace: claude plugin marketplace add getsentry/skills.
  2. Install the plugin: claude plugin install sentry-skills@sentry-skills.
  3. Restart Claude Code; skills activate automatically when relevant. Update later with claude plugin marketplace update then claude plugin update sentry-skills@sentry-skills (or via the /plugin manager).
  4. Alternatively, for any skills.sh-compatible agent (Cursor, Cline, GitHub Copilot, and others): npx skills add getsentry/skills.

Prerequisites

Claude Code (or another supported harness); Node.js for the npx installer.

Use cases

codetestingdevops

Works with

Claude CodeCursorCopilotAny agent (.md)

Trust score — 7/10

/10 trustMetrics-checked

Official first-party collection from the Sentry org (Apache-2.0), following the Agent Skills open format, with genuinely useful security-oriented skills (skill-scanner, secret-serialization leak detection, GHA security review). Reservations: many skills are Sentry-internal (brand guidelines, Django access/perf review, SRED tax-credit tooling) with limited value outside Sentry; repo is young (created Jan 2026, 1k stars); independent coverage is thin — adoption evidence is mostly forks of the same README. Metrics-v1 score 7/10 — adoption 6, maintenance 6.9, openness 10, manual 7.

★ 1,047 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
6.0/10
Maintenance · recency, releases, issue load
6.9/10
Openness · license + docs
10.0/10
Manual review · sanity read + red-flag check
7.0/10
  • · 1,047 GitHub stars (log-scaled vs catalog peers)
  • · last push 3d ago; 0 releases/12mo; 27 open issues
  • · manual review: several skills Sentry-internal (limited outside value); thin independent coverage

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.