gh-fix-ci

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and ingests GitHub Actions run and job logs (via gh run view --log and gh api /repos/<owner>/<repo>/actions/jobs/<job_id>/logs in scripts/inspect_pr_checks.py and the SKILL.md workflow), treats those user-generated CI logs as the source for failure snippets, and uses them to summarize causes and drive fix plans, so untrusted third-party content can influence agent actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill invokes the GitHub CLI at runtime to fetch Actions run and job logs (e.g., https://github.com///actions/runs/<run_id> and the API endpoint /repos/{owner}/{repo}/actions/jobs/{job_id}/logs), and those fetched logs are extracted and injected into the agent's summaries/snippets used to generate fixes — meaning external content fetched at runtime can directly influence prompts/instructions.