The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown instructions and documentation for code refactoring best practices. No malicious commands, unauthorized file access, or credential exposure patterns were found.
[EXTERNAL_DOWNLOADS]: The skill references guidelines from a trusted organization's GitHub repository for inspiration and context. This is a static reference to an official source and does not involve automated or suspicious downloads.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided code, which is an inherent attack surface for indirect prompt injection. The skill addresses this by establishing clear process boundaries: it mandates following project-specific conventions (e.g., CLAUDE.md), requires exact behavioral preservation, and dictates verification through the project's own test suites and linters. These constraints serve as safety measures for the agent's operations on untrusted data.

code-simplification