The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use local shell commands for development tasks, including word counting (wc -w) for token optimization and executing a local utility script (./render-graphs.js) to generate SVG visualizations from Graphviz diagrams.
[EXTERNAL_DOWNLOADS]: The documentation references agentskills.io/specification as the authoritative source for YAML frontmatter fields and skill structure standards. This is a legitimate technical reference for skill development.
[PROMPT_INJECTION]: The file contains a section on 'Bulletproofing Skills Against Rationalization,' which teaches techniques for making instructions more authoritative and resistant to agent 'loopholes.' These include using persuasion principles (e.g., Cialdini's principles) and explicit negative constraints. While these are behavioral modification techniques, they are used here as a quality assurance methodology to enforce technical discipline (like TDD) rather than to bypass safety protocols.

writing-skills