The Agent Skills Directory

[SAFE]: The skill provides comprehensive and secure engineering guidelines for refactoring tasks, emphasizing behavior preservation through equivalence checks and manual verification of deletions.
[SAFE]: No evidence of credential harvesting, data exfiltration, or unauthorized command execution was found. The instructions rely on project-local configurations and standard development workflows.
[SAFE]: While the skill involves processing source code which constitutes an indirect prompt injection surface, this activity is inherent to its primary purpose as a development tool. The instructions for manual review and output verification act as safeguards against automated exploitation.

write-refactor