content-strategy
Warn
Audited by Snyk on Feb 28, 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). SKILL.md explicitly instructs the agent to use web search and scrape public, user-generated sources (e.g., "Forum Research" citing Reddit, Quora, Hacker News, Product Hunt and competitor sites via site:search) and to extract/act on that material for content decisions, which exposes the agent to untrusted third-party content that could carry indirect prompt injections.
Audit Metadata