the-lead-router
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill incorporates robust instructions to mitigate indirect prompt injection. Through the included security reference (references/agent-security.md), it explicitly directs the agent to treat all ingested content as untrusted data and forbids following instructions embedded within such content.
- [SAFE]: The skill is entirely instructional and logic-based, containing no executable code, tool definitions, or network capabilities. It defines an evidence chain for processing data which includes ingestion points like product context files and user input, bounded by clear rules to ignore embedded instructions and requirements for HTML-escaping any generated markup.
Audit Metadata