feedback-analysis
Warn
Audited by Snyk on Mar 13, 2026
Risk Level: MEDIUM
Full Analysis
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 aggregating and analyzing untrusted, user-generated content from public channels (e.g., "Multi-channel collection" lists app store reviews, social media mentions like Twitter/Reddit/HN, and community forums such as Discord/Slack/GitHub Issues), which the agent is expected to read and use to drive classification, prioritization, and product decisions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata