sales-octolens
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: 0.95). Outsider-authored free text is ingested into the agent’s LLM context via the runtime mention payload: the skill calls Octolens
/mentions(or receives webhook/MCP results) where each mention includescontent(“Full text of the mention...”) authored by third parties (e.g., Reddit/Twitter/etc.).
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata