agent-orchestrator
Warn
Audited by Snyk on Feb 15, 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's Research Agent template (references/sub-agent-templates.md) explicitly gives agents WebSearch/WebFetch tools and instructs them to "gather information from web searches" and "retrieve specific web pages" (with outputs including sources.md), meaning agents will fetch and read untrusted public web content.
Audit Metadata