Publisher: Sentry · Editorial profile by AI Tools Ratings · Review method
What task is this skill designed for?
Inspect GitHub Actions workflows for a concrete path by which an attacker without repository write access could influence privileged execution. Useful for repositories accepting external contributions.
What input and output should you expect?
Input: Provide workflow YAML, local composite actions, scripts and configuration referenced by those workflows. Specify triggers, token permissions and which inputs can come from a fork.
Output: An attack-path-oriented static review connecting attacker control, workflow behavior and potential privilege impact, with specific mitigation suggestions.
What does it need to run?
Read access to .github/workflows and referenced local files. The source includes topical references for expressions, credentials, runners and related CI risks.
Publisher documents Claude Code installation and skills.sh support for compatible agents including Cursor, Cline and GitHub Copilot. Host permissions and invocation behavior still require validation.
How should you approach installation?
- Read the pinned SKILL.md and repository LICENSE; compare project-specific conventions with your own instructions.
- Follow the publisher installation guidance for your host, retaining the full skill folder and references; begin in a disposable repository with minimal permissions.
- Run the suggested draft or read-only trial, review its acceptance checklist, and grant write or remote permissions only for the intended task.
Installation guide: discovery, dependencies and runtime checks
What files, accounts or data can it access?
Repository reading is sufficient for an initial review. Do not run an exploit in live CI, expose secrets or open an attacker PR as part of a casual trial; source context may reach the host model.
What does the skill cost?
No separate Skill purchase price is stated in the reviewed repository. The host agent subscription or model usage is separate; repository services and any authorized CI jobs may add costs. A local drafting trial does not require paid remote execution.
What are its limits and alternatives?
The source has a specific fork-attacker threat model and does not comprehensively audit every referenced external repository. Absence of findings is not a supply-chain certification; broader pinning or infrastructure policy may require a separate review.
Use a workflow linter for syntax and a dedicated CI security audit for organizational runners, external actions and privileged infrastructure.
What could you ask in a small trial?
Try this with a small sample:
Statically inspect this synthetic pull_request_target workflow that checks out a fork commit and runs its build script. Identify the trust transition and available token scope. Suggest a safer separation of jobs; do not trigger Actions or create a proof-of-concept PR.
How would you verify the result?
- The finding names the untrusted input, execution step and privilege boundary.
- Assumptions about trigger behavior and token permissions are explicit.
- No live workflow, secret read or external write is performed for the trial.
Use a baseline and record acceptance evidence before treating a successful task as a reliable workflow.
Which source and license were checked?
Read the reviewed SKILL.md at its pinned commit · Complete reviewed folder
Repository and current publisher guidance
Checked:
Apache-2.0 — repository LICENSE at the reviewed commit. License evidence
Version details and review scope
Source: getsentry/skills/skills/gha-security-review
Reviewed commit: d12f250460ffd961f549dbebf7eeb0cf4afd4c2e
SKILL.md SHA-256: b75c6c43ff99c82b3cf44519bd30723917de97064f225e392ecfd6fdcd7455bd
Source-document review. This profile covers the source workflow, stated dependencies and license evidence. See our editorial methodology for how we distinguish source reviews and hands-on evaluations.