Publisher: OpenAI · Editorial profile by AI Tools Ratings · Review method
Not runtime tested; source-document review only. No security certification.
What task is this skill designed for?
Investigate a failing pull-request check and produce a proposed fix tied to the relevant GitHub Actions failure.
What input and output should you expect?
Input: A repository checkout, a pull-request reference and an authenticated GitHub CLI session with the access needed to inspect its checks.
Output: Expected, not independently tested: a failure summary, supporting log context and a repair plan; code changes follow user approval in the source workflow.
What does it need to run?
Codex command and repository access, GitHub CLI authentication and Python for the bundled inspection helper. Confirm repository access with the account owner; do not grant broader token scopes merely to follow a generic example.
Author-provided Codex skill using GitHub CLI; other agents have not been verified.
How should you approach installation?
- Read the pinned source and bundled inspection helper before providing repository access.
- Install using the repository's current Codex guidance; installation is not independently verified here.
- Use a test repository or non-sensitive failure first. Review the diagnosis and approve a concrete plan before allowing changes.
Installation guide: discovery, dependencies and runtime checks
What files, accounts or data can it access?
Reads repository and CI metadata through authenticated GitHub access. Approved fixes can modify working files. CI logs can contain sensitive information, so review excerpts before sharing them externally.
What does the skill cost?
No skill purchase is listed. GitHub account features and agent usage have their own costs.
What are its limits and alternatives?
The source workflow supports GitHub Actions investigation and excludes repairs to external providers. Missing or expired logs can prevent a diagnosis. A local repair does not establish that remote checks passed.
Inspect checks directly with GitHub CLI for a single known failure. This skill is more useful when an agent needs a repeatable diagnosis, plan and approval sequence.
What could you ask in a small trial?
Try this with a small sample:
Investigate only the failing GitHub Actions check on this test PR. Cite the relevant log evidence, propose a repair, and wait for approval before editing files or posting comments.
How would you verify the result?
- Check the failure belongs to the correct pull request and run.
- Map the proposed fix to an actual log or code condition.
- After an approved change, distinguish local test results from the remote CI rerun.
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 — identified in skills/.curated/gh-fix-ci/LICENSE.txt at the reviewed commit. Third-party assets and dependencies may have separate terms. License evidence
Version details and review scope
Source: openai/skills/skills/.curated/gh-fix-ci
Reviewed commit: 49f948faa9258a0c61caceaf225e179651397431
SKILL.md SHA-256: 7b326b4a2f0f5f85122144628ec02077e48841e0e0e82efce88b3415bcfb7c26
Source-document review. Source text and stated dependencies were inspected for this profile. Referenced code was not comprehensively audited, no runtime test was performed, and source files are not mirrored here.