The Agent Skills Directory

[SAFE]: The skill implements a defensive development workflow that prioritizes isolation and user confirmation. It explicitly requires creating dedicated branches and prohibits destructive actions like branch deletion or worktree cleanup without confirmation.
[SAFE]: The instructions include robust verification rules, such as requiring production fixes over test modifications and ensuring automated tests are not ignored. This prevents the silent degradation of the codebase's integrity.
[SAFE]: No obfuscated content, hardcoded credentials, or suspicious network operations were identified. The workflow relies on standard local repository operations and well-defined tool interactions.

task-execution-simple