The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a utility script render-graphs.js that uses child_process.execSync to execute the system dot command (from Graphviz). This script renders SVG diagrams from markdown code blocks found within the skill's documentation.
[PROMPT_INJECTION]: The skill provides detailed instructions on using persuasion principles (Authority, Commitment, Scarcity) and strong imperative language ("YOU MUST", "No exceptions", "Delete means delete") to reinforce agent compliance with specific procedural rules. This approach is intended to prevent agents from rationalizing their way out of following complex or high-effort workflows like TDD.
[DATA_EXPOSURE]: The utility script render-graphs.js performs filesystem operations by reading content from SKILL.md files and writing rendered SVG output to a local diagrams directory.
[SAFE]: References to external resources and guidelines (such as Anthropic's best practices) point to well-known domains and official documentation, and are documented neutrally within the skill's instructions.

writing-skills