util-research-library
Warn
Audited by Snyk on Feb 24, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 1.00). The SKILL.md workflow explicitly requires performing WebSearch and ingesting content from public, user-generated sources (e.g., reddit.com, news.ycombinator.com, github.com, stackoverflow.com, NVD/Snyk) during Phase 1 and using that evidence to score dimensions and drive ADOPT/MIGRATE/STAY/AVOID recommendations, so untrusted third‑party content can directly influence agent decisions.
Audit Metadata