The Agent Skills Directory

[SAFE]: The skill implements a robust decision-making framework for technical validation. It includes explicit instructions to protect sensitive information, such as the mandate to use synthetic or anonymized data rather than production data, and to keep credentials within approved local mechanisms. All analyzed files (SKILL.md, templates, and evaluations) focus on legitimate developer workflows like benchmarking, integration probing, and feasibility testing. The routing logic correctly distinguishes between experimentation and production delivery, reducing the risk of accidental deployment of unstable code. No external network execution, obfuscated content, or unauthorized privilege escalation patterns were found.

feature-spike