sales-kwatch
Warn
Audited by Snyk on Jun 14, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.85). KWatch’s runtime workflow sends outsider-authored mention text (e.g., Reddit/HN post/comment body) to the agent via the configured
api_webhook_urlas the JSON fieldcontent, which is then ingested into the agent’s LLM context (indirect prompt injection risk).
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata