lamina-onboarding

# Onboarding (agent-native) Onboarding = **primary actor completes first valuable workflow** without documentation — verified by actor walk, not tutorial copy alone. ## Contract encoding - First-run workflow flagged in `workflows[]` (e.g. `onboarding: true`) - `empty_states` + `surfaces[]` for first dashboard visit - Power features in `secondary_actions` / progressive disclosure notes - Optional `scenarios[]` for skipped onboarding — actor still reaches goal ## Frameworks - **Working set**: optimize discoverability for daily operations after day one. - **Command modalities**: pedagogic (visible labels) → immediate (shortcuts) — spec accelerators in `implement.md`, not required for pass. - **Idioms over metaphors**: consistent patterns across screens — reference in `decisions.md`. ## Design checklists 1. First productive action obvious on empty state (`empty-states`). 2. No mandatory feature tour blocking primary workflow. 3. Help/docs are fallback — interface teaches via signifiers. 4. Novice path ⊆ expert path (experts aren't forced through wizards). 5. Onboarding workflow has full scenario coverage. ## Verify checks - Fresh-actor subagent (low `technical_literacy` persona if defined) completes first workflow. - No dead ends before first value delivered. ## Anti-patterns - Multi-step modal tour before any action. - Expert-only shortcuts with no discovery path. - Metaphor-bound IA that breaks at scale. ## Related - [Empty States](../lamina-empty-states/SKILL.md) ...