review-skill

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

• Third-party content exposure detected (high risk: 0.75). The workflow reads SKILL.md and “all files in the skill directory” for each changed skill (Step 3), and those files are authored by other contributors/outsiders in the repo/PR context, so their free-text content can be ingested into the agent’s LLM context during review.