Agents
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 explicitly instructs spawned agents to perform live web and social-media searches and ingest third-party content (e.g., PerplexityResearcherContext.md describes real-time web search via the Perplexity API, CodexResearcherContext.md shows codex exec --sandbox danger-full-access "research query" for live web access, and GrokResearcherContext.md references real-time X/Twitter analysis), so untrusted public content is fetched and used to drive agent decisions.
Audit Metadata