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
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 26, 2026, 04:17 AM
Issues
1
Security Audit — snyk — social