expert-instruction

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The documentation includes common prompt injection strings such as "Ignore previous instructions" and "Reveal your system prompt". These are utilized exclusively as examples for adversarial testing and instruction design, serving an educational purpose rather than representing a malicious attempt to subvert the agent.\n- [EXTERNAL_DOWNLOADS]: The skill references a command to install an additional module from the vendor's repository using pnpm dlx skills add oakoss/agent-skills. This represents standard functionality for extending skill capabilities within the author's ecosystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 11:51 AM
Security Audit — agent-trust-hub — expert-instruction