Sentry Agent Skills - Setup Guide¶
Source: getsentry/skills GitHub: getsentry/skills Skills: 31 skills · 85.3K total installs Category: Code Review & Security First Seen: catalogued August 15, 2026 evening sweep Quality Tier: 🟢 Production (official org - Sentry, the application monitoring vendor)
Sentry's official agent skill suite encodes the review and release hygiene its own engineers use: security review against OWASP-class issues, code simplification, bug finding, PR writing and iteration, GitHub Actions security review, and internal ops skills (agents-md, claude-settings-audit, sred work summaries). Queued in prior sweeps at a 20.7K API-sum estimate; the publisher page shows 85.3K - a 4x jump, which is why it leads this batch.
Installation¶
npx skills add getsentry/skills
Prerequisites¶
| Requirement | Details |
|---|---|
| Node.js + npx | For the skill installer |
GitHub CLI (gh) |
For gha-security-review, gh-review-requests, iterate-pr, pr-link-issue |
| Django project | For django-perf-review and django-access-review only |
What It Provides¶
| Skill | Installs | Purpose |
|---|---|---|
| security-review | 13.8K | Full security review of code changes |
| code-simplifier | 7.4K | Simplify code without behavior changes |
| find-bugs | 5.0K | Hunt defects in changed code |
| agents-md | 4.8K | Maintain AGENTS.md guidance files |
| code-review | 4.4K | Systematic code review workflow |
| commit | 4.0K | Conventional commit messages |
| gha-security-review | 3.8K | Audit GitHub Actions workflows |
| iterate-pr | 3.8K | Iterate on PR feedback |
| claude-settings-audit | 3.8K | Audit agent settings files |
| pr-writer | 3.8K | Draft PR descriptions |
| create-branch / gh-review-requests | 3.6K each | Branching and review-request management |
| skill-scanner | 2.6K | Scan installed skills |
| blog-writing-guide | 2.3K | Sentry-style blog drafts |
| presentation-creator | 2.0K | Build slide decks |
| django-perf-review / django-access-review | 1.7K / 1.5K | Django performance and access control review |
| skill-writer / skill-creator | 1.7K / 426 | Author new skills |
| brand-guidelines | 1.6K | Sentry brand voice |
| doc-coauthoring | 1.5K | Collaborative documentation |
| sred-work-summary / sred-project-organizer | 1.4K each | SRE duty summaries and project tracking |
| prompt-optimizer | 1.3K | Optimize agent prompts |
| replay-ux-research | 816 | UX research via session replay |
| create-pr / pr-link-issue / document-api-endpoint | 633 / 586 / 495 | PR and API documentation helpers |
| triage-frontend-issues | 587 | Frontend issue triage |
| typing-exclusion-worker / warden-lint-judge | 910 / 37 | Lint and typing support |
Quick Start¶
- Install:
npx skills add getsentry/skills - Start with
security-revieworcode-reviewbefore merging any PR - Ask: "run a security review on my changes and draft the PR description"
CorpusIQ Use Cases¶
| Use Case | How |
|---|---|
| PR security gate | security-review + gha-security-review as an agent-side review layer for our GitHub work |
| Repo hygiene | code-simplifier and find-bugs for maintenance passes on internal tooling |
| PR automation | pr-writer, iterate-pr, and gh-review-requests replace manual PR boilerplate |
| Sentry integration pairing | Natural fit when instrumenting projects with Sentry error monitoring |
| Reference skills | skill-writer and skill-creator as templates for our own skill packaging |
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.
Limitations / Verification¶
- Publisher-page install counts verified; individual skill audit pages were not fetched for this multi-skill suite
- Django-specific skills require a Django project to be useful
warden-lint-judge(37 installs) is effectively brand-new - treat as untested
npx skills add getsentry/skills # verify install works
Related¶
Powered by CorpusIQ