The Agent Skills Directory

[PROMPT_INJECTION]: The skill does not contain any instructions that attempt to override agent behavior, bypass safety filters, or extract system prompts. The 'Avoiding AI-Generated Language' section is focused purely on stylistic and tonal improvements.
[DATA_EXFILTRATION]: No patterns for accessing sensitive local files (~/.ssh, ~/.aws, etc.) or exfiltrating data to external domains were detected. All network references are to reputable documentation sources (MDN, javascript.info).
[REMOTE_CODE_EXECUTION]: There are no commands or instructions that download and execute remote scripts (e.g., curl|bash). References to testing frameworks like Vitest are provided as templates for developer workflow rather than runtime execution.
[EXTERNAL_DOWNLOADS]: The skill does not perform any external package installations (npm, pip) or fetch unverified dependencies. All external URLs point to well-known documentation and educational services.
[COMMAND_EXECUTION]: The skill does not instruct the agent to execute dangerous shell commands or acquire elevated privileges (sudo/chmod). It focuses on generating Markdown and code examples for documentation.

write-concept