review-pr

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and reads GitHub PR diffs and full file contents (via gh pr view, gh pr diff, and the Read tool), which are user-generated public content the agent must interpret to decide and post reviews, enabling indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches PR metadata and full file contents at runtime (via gh pr view / gh pr diff and uses file URLs like https://github.com/{owner}/{repo}/blob/{headRefOid}/{file_path}), and that fetched content is injected into the agent's context to drive prompts and generate the review, so it directly controls agent behavior and is a required runtime dependency.