golang-continuous-integration

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow for the AI code review (assets/claude-code-review.yml) runs on pull request events and has the LLM read untrusted outsider-authored free text at runtime via gh pr diff, gh pr view, and the review/comment thread (including pull_request_review_comment / pull_request_review payloads) which are attacker-controlled PR description/diff and comment text.

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

• Potentially malicious external URL detected (high risk: 0.90). The GitHub Actions workflow runs the command "npx skills add https://github.com/samber/cc-skills-golang ..." at runtime which fetches remote skill files that are loaded by the AI reviewer and directly control its prompts/instructions (https://github.com/samber/cc-skills-golang).