youtube

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly shows an API key example (AIzaSy...) and uses that key as a literal/query parameter in curl commands (key=$YOUTUBE_TOKEN), which encourages copying a real secret into commands the agent may output verbatim, creating an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the agent to fetch and read user-generated YouTube content (e.g., "Get Video Comments" and "Get Video Details"/search endpoints that return video descriptions and comments via the YouTube Data API), so untrusted third-party text could be interpreted and influence the agent's actions.