The Agent Skills Directory

[SAFE]: The skill facilitates the creation of a doc-as-code environment by installing Markdown templates and utility scripts. All executable content, such as the tool skeletons (tool-*.mjs) and the change management script (changes.mjs), is sourced from local templates within the skill package.
[SAFE]: The design of the operational tools (tool-*.mjs) demonstrates a security-first approach by using command-line flags for sensitive connection parameters (hosts, tokens, passwords) and including specific sanitization logic to mask credentials and truncate large data payloads in the agent's output.
[SAFE]: The skill enforces architectural constraints that require the agent to use standardized, auditable tool interfaces for all system interactions, preventing bypasses or direct access to sensitive infrastructure like databases or external APIs without proper encapsulation.

project-scaffold