The Agent Skills Directory

[SAFE]: The skill's primary purpose is managing local markdown files within the .claude/rules/ directory structure, which is a standard configuration path for Claude Code.
[COMMAND_EXECUTION]: The skill utilizes local Python scripts (list_rules.py and validate_frontmatter.py) for administrative tasks. Analysis of these scripts confirms they only perform local file I/O within expected paths and use yaml.safe_load() to prevent code execution during configuration parsing.
[DATA_EXFILTRATION]: No network-capable code or exfiltration patterns were detected. The scripts do not use libraries like requests or urllib, and no curl or wget commands are present in the skill instructions.
[PROMPT_INJECTION]: The skill instructions do not contain any patterns designed to override agent safety filters or ignore system prompts. It uses standard AskUserQuestion patterns for eliciting user preferences.

rules-builder