feature-video

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly reads PR metadata and the PR body from GitHub (e.g., gh pr view [number] --json body -q '.body') and uses that user-generated PR content to plan and guide the walkthrough, so it ingests untrusted third-party input that could carry indirect prompt-injection content.