The Agent Skills Directory

[SAFE]: The skill consists of instructional markdown content designed to improve code readability and maintainability. All code snippets provided are benign examples of standard refactoring techniques like guard clauses, mapping objects, and early returns.
[PROMPT_INJECTION]: There are no instructions that attempt to override agent safety protocols, bypass filters, or extract system prompts. The language is purely instructional and focused on code quality.
[DATA_EXFILTRATION]: No commands for accessing sensitive files (e.g., credentials, .env, SSH keys) or performing unauthorized network operations were found. References to project-specific files like CLAUDE.md are standard for AI coding agents.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute any remote scripts. It mentions the use of standard development tools like git blame, linters, and testing suites, which are appropriate for its stated purpose.
[COMMAND_EXECUTION]: References to shell tools such as sed or AST transforms are framed as suggestions for large-scale refactors (The Rule of 500) and do not include any immediate or dangerous automated execution patterns.

code-simplification