nerd-patrol
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external content (code and configurations), which inherently introduces an attack surface for indirect instructions.
- Ingestion points: The skill body in
SKILL.mdspecifies the examination of "confirmed code, pull request, module, application, endpoint, dependency, or configuration scope." - Boundary markers: The skill relies on structured templates (Patrol Scope, Security Finding) but does not define explicit delimiters or instructions to ignore embedded prompts in audited data.
- Capability inventory: Instructions mention an "execute endpoint" and permit the agent to "change code," suggesting the presence of file system write or shell execution capabilities.
- Sanitization: The skill implements safety measures by requiring the use of "harmless" payloads and "sanitized" requests, which serves to mitigate the risk of unintended code execution or exfiltration.
Audit Metadata