The Agent Skills Directory

[SAFE]: The skill is primarily educational and organizational, providing guidelines for skill authoring. It encourages the use of Test-Driven Development for documentation to ensure reliability and compliance with project standards.
[COMMAND_EXECUTION]: The render-graphs.js script uses child_process.execSync to run the system dot command (from Graphviz). This is a local development utility intended for rendering Graphviz diagrams found in skill documentation into SVG format. It correctly passes input via stdin rather than command-line arguments, which is a secure implementation for this use case.
[EXTERNAL_DOWNLOADS]: The documentation references official Anthropic best practices and provides instructions for installing Graphviz via well-known and trusted system package managers (Homebrew on macOS and APT on Linux). No suspicious or unverified external downloads were detected.
[PROMPT_INJECTION]: The skill focuses on enforcing discipline and strictly following rules (e.g., the 'Iron Law' of TDD). It does not contain instructions that attempt to bypass safety filters or override system prompts in a malicious way; instead, it provides techniques to make agent behavior more predictable and compliant.

writing-skills