gh-actions-debug

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

• Third-party content exposure detected (medium risk: 0.30). Source is GitHub Actions run/job data that the workflow fetches via gh run view / gh api (including job.log and derived errors.txt), where the text is ultimately sourced from repository-owner workflow logs and step outputs, which an outsider can influence by pushing commits that run those workflows.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime scripts call GitHub API endpoints to fetch repository content that is then read into the agent for analysis (e.g., "repos/$REPO_REF/contents/$WPATH" and "repos/$REPO_REF/actions/runs/$RUN_ID/jobs"), meaning external content fetched at runtime can directly control the agent's prompts/analysis.