cro-methodology
Warn
Audited by Snyk on Feb 19, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). Yes — SKILL.md (see "Gather Market Intelligence") and the referenced docs (OBJECTIONS.md, RESEARCH.md) explicitly instruct monitoring and analyzing social media, public reviews and forums (e.g., "Reddit, Twitter threads", "Reviews") — i.e., untrusted, user-generated third‑party content the agent would fetch/ingest and that can materially influence its CRO decisions and next actions.
Audit Metadata