Agent Development

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely composed of educational markdown files, agent configuration templates, and a local validation script.
  • [COMMAND_EXECUTION]: The script 'scripts/validate-agent.sh' is a local utility that uses standard text-processing commands (sed, awk, grep) to perform static analysis on user-provided agent files. It does not execute code from external sources or the internet.
  • [PROMPT_INJECTION]: The provided agent templates and system prompt guidelines focus on establishing clear expert personas and behavioral boundaries. No bypass patterns or instructions to ignore safety protocols were detected.
  • [DATA_EXFILTRATION]: No network-enabled tools, hardcoded secrets, or patterns suggesting the collection of sensitive local data are present in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 07:31 PM