social
Warn
Audited by Snyk on Sep 11, 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 skill includes social listening workflows that instruct the agent to fetch external posts from Reddit, Hacker News, Bluesky, and RSS feeds using
curlrecipes, allowing outsider-authored free text to enter the agent's runtime context.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata