The Agent Skills Directory

[SAFE]: The skill is a collection of educational resources, templates, and architectural guides. A thorough audit of the 21 files reveals no malicious patterns, credential harvesting, or unauthorized data exfiltration.
[COMMAND_EXECUTION]: The utility script render-graphs.js uses child_process.execSync to invoke the local Graphviz dot utility for rendering diagrams. This tool is designed for local development use and does not execute automatically or process untrusted remote inputs.
[PROMPT_INJECTION]: The skill includes documentation on 'persuasion' and 'anti-rationalization' principles. These are meta-prompting techniques aimed at preventing agents from skipping required steps in discipline-heavy workflows (such as TDD). These instructions are educational and do not target the bypass of platform safety filters or core system prompts.

writing-skills