refactoring-agent-instructions
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard shell commands to create a directory structure (mkdir -p guidelines) for organizing refactored instructions. This operation is limited to the project root and is essential for the skill's functionality.
- [DATA_EXPOSURE]: The skill reads content from specific instruction-related files like CLAUDE.md, .cursorrules, and .claude/settings.json to extract and reorganize directives. This access is localized to files identified through its discovery patterns.
- [SAFE]: The analysis found no evidence of prompt injection, data exfiltration, or obfuscation. The workflow is designed with multiple verification checkpoints where the user must confirm file discovery, contradiction resolution, and final file deletions, ensuring the agent operates with explicit consent.
Audit Metadata