building-claude-code-skills

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs the agent to fetch and read user-generated content from public GitHub (e.g., "gh pr view", "gh pr diff", "gh issue view", and gh API commands) as part of review workflows, which are open/public third‑party sources and thus expose the agent to untrusted content that could enable indirect prompt injection.