sales-keymentions
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: 0.75). KeyMentions’ runtime workflow can ingest outsider-authored Reddit thread text (public posts/comments) into the agent’s LLM context to generate or auto-publish AI comments, creating an indirect prompt-injection surface from that untrusted content.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata