sales-prems
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). Outsider-authored free text is ingested into the agent’s LLM context via Prems webhook payloads (e.g.,
content,title,ai_pitch) that originate from monitored third-party platforms like Reddit/X/LinkedIn, which Prems then forwards to the agent at runtime.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata