Audience: government academy course teams · Artifacts: builder.html (2D) · 3d.html + builder.js (3D module track) · shared data scenes-data.js · Case study: index.html
In one line: a lesson-scene builder where every placed scene is scored on one lens — does the interaction teach (retrieval, decisions, reflection) or decorate (click-next polish, hotspot theatre)?
Classroom days get converted to e-learning under deadline pressure, and the easy path — polished click-next slides — ships on time and teaches nothing. The builder reproduces the real constraint: an effort budget forces the trade-off between engagement decoration and learning interaction, with instant verdicts on each placement.
Three module briefs (incident refresher, intake conversation skills, data classification) each with a required scene type and its own rule function. Eight scene cards cost 0-3 effort points; the engagement budget bar forces trade-offs. Every placement gets a TEACHES/DECORATES verdict; the quality check scores the build, flags budget overruns, and names the meta-skill.
Scoring rules are design heuristics, not empirical laws. The briefs are composites of government-academy training patterns, not a real agency's curriculum.
Part of I Wayan Usada's learning-design portfolio · all clients and scenarios are fictional composites built from realistic domain patterns · drafted with AI assistance and fully rewritten, fact-checked and owned by a human.
action-map.md — Action map — the behaviors the artifact must produce, traced to the business goalstoryboard.md — Text-based storyboard — the scene plan written before buildvisual-mockups.md — Visual mockups & style guide — mood, type, key screens, component decisions