gtm-partner

Warn

Audited by Snyk on Feb 22, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.90). The skill's Phase 0/Phase 1 workflow explicitly reads ~/.idea-validator/latest-evaluation.json which contains market_research_raw drawn from public sources (Reddit, competitors, Google Trends, YouTube) and requires the agent to use those findings to pre-fill audience, cite pain points, and drive channel/asset-generation decisions, so it consumes untrusted third-party content that can influence actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 22, 2026, 08:57 PM