The Agent Skills Directory

[SAFE]: The skill serves as a high-level orchestrator for a multi-stage pipeline designed to build interactive teaching websites. It coordinates specialized sub-skills and enforces structural gates to ensure that prerequisites, such as outlines and unit content, are completed before progressing to code generation or publishing.
[COMMAND_EXECUTION]: The skill instructions mention standard development tools and commands, including npm run serve, git log, Playwright for automated testing, and Pandoc for document conversion. These are legitimate tools for the project's stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where user-supplied content is transformed into executable code and document artifacts.
Ingestion points: Project markdown files (overview.md, outline.md, content.md), CSV/YAML data, and user chat prompts.
Boundary markers: None explicitly defined in the orchestrator instructions to separate data from instructions during the transformation process.
Capability inventory: Sub-skills perform file system writes (SPA generation), execute shell commands (npm, git), and invoke external utilities (Playwright, Pandoc).
Sanitization: Not explicitly implemented in the instructions; the skill relies on human review and automated verification sub-skills (web-content-audit, web-visual-verification) for output integrity.

teaching-site