tech-decision
Warn
Audited by Snyk on Mar 5, 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). This skill's SKILL.md Phase 2 explicitly calls the dev-scan skill to fetch community content from public, user-generated sites (Reddit, HN, Dev.to, Lobsters), which agents are required to read and incorporate into the tradeoff-analyzer and decision-synthesizer workflow, creating a clear vector for indirect prompt injection.
Audit Metadata