The Agent Skills Directory

[SAFE]: The skill is entirely instructional and focuses on prompt engineering best practices. No malicious patterns, obfuscation, or attempts to bypass safety guidelines were found.
[COMMAND_EXECUTION]: The skill identifies several tools such as Read, Write, Edit, Bash, and WebSearch as part of its recommended refactoring and research workflow. These references are appropriate for the functional purpose of managing and updating local project files.
[DATA_EXFILTRATION]: While the skill mentions network-capable tools like WebSearch and WebFetch, they are explicitly designated for documentation research and verification. No evidence of unauthorized data transfer or access to sensitive local files was detected.

subagent-refactoring-methodology