メニュー
Core pedagogy for Hi Tuto lessons — one mechanism per lesson, prerequisite ordering, active recall. Use when scoping or sequencing lesson content. (Body pending Deep Agents skills wiring.)
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Hi-Tuto design system for all frontend/UI work in this repo. Use whenever creating or modifying React components, pages, styles, Tailwind classes, or the mascot in frontend/ — ensures the Thinkby-inspired look (flat vivid color panels, one-sided shadows, square mascot) stays consistent.
Author Hi Tuto HTML lesson capsules. Use when emitting final lesson HTML, choosing scroll vs Canvas Studio vs slide shell, wiring tutor hooks, or applying postprocess/CSP constraints.
Source-grounding rules — cite retrieved facts, never invent specifics, prefer source packs over web when a document is attached. Use when a lesson has source packs or research. (Body pending Deep Agents skills wiring.)
Data-synthesis guidance — generating plausible datasets for visualizations or simulations when no source data exists. Use during asset_plan for data-driven capsules. (Body pending Deep Agents skills wiring.)
Generate PBR meshes (Hunyuan 3D Pro) for Canvas Studio chapters only. Use when presentation is studio and subjects need realistic 3D. Studio chrome lives in ui-studio-style — this skill owns mesh generation only.
Simulation guidance — parameter sliders, physics/agent loops, and deterministic seeds. Use for simulation- or tool-archetype lessons. (Body pending Deep Agents skills wiring.)
| name | pedagogy |
| description | Core pedagogy for Hi Tuto lessons — one mechanism per lesson, prerequisite ordering, active recall. Use when scoping or sequencing lesson content. (Body pending Deep Agents skills wiring.) |
Placeholder. Progressive disclosure via frontmatter description when Deep Agents skills wiring lands.
