social
Warn
Audited by Snyk on Aug 26, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The “social” skill’s listening/engagement triage workflow explicitly pulls in outsider-authored free text at runtime by fetching Reddit/HN/Bluesky posts (e.g.,
selftext/comment text) and extracting page text from LinkedIn/X feeds in order to score and draft replies.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata