skill-optimizer

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill package is composed exclusively of Markdown files providing methodological guidance for prompt engineering and benchmarking. It does not contain any executable scripts, binaries, or automated tasks.- [NO_CODE]: No Python, Node.js, or shell scripts are included. The skill functions as a set of 'best practice' instructions for the agent to follow when assisting with skill development.- [DATA_EXPOSURE]: No hardcoded credentials, API keys, or access to sensitive file paths were detected. The skill does not perform any data exfiltration or network communication.- [PROMPT_INJECTION]: The instructions are meta-instructional, aimed at improving the 'salience' and 'activation' of other skills. They do not contain techniques to bypass safety protocols or extract system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 02:03 PM