Skill map

Lamina includes ~60 capability skills organized into three taxonomies. The orchestrator routes design and verify sections to the right skill via audit profiles.

Quick path

Focused question? /lamina <question> routes to one skill (direct mode)
Full workflow? /lamina-init first, then /lamina-design or /lamina-verify loads multiple skills via audit profiles
Deep dive? Each skill has a SKILL.md in the Lamina repo

Systems thinking

The structural spine — entities, rules, dependencies, and traps.

Problem signal	Skill
Entities, relationships, purpose	`lamina-system-structure`
Feedback, delays, oscillation	`lamina-feedback-loops`
Fixes keep failing, structural traps	`lamina-system-traps`
Where to intervene (rules vs UI)	`lamina-leverage-points`
Impossible states, business rules	`lamina-invariants`
Feature reachability, unmet prerequisites	`lamina-dependencies`
Domain boundaries, hide complexity	`lamina-modularity-boundaries`
What users see when (stale data)	`lamina-consistency-guarantees`
Double-submit, concurrent edits	`lamina-idempotency-concurrency`
Name trade-offs before mechanisms	`lamina-tradeoffs`
Multi-actor views stay consistent	`lamina-multi-view-integrity`
Evolving rules safely	`lamina-evolutionary-rules`
Notifications, downstream updates	`lamina-side-effects`

UX and product expression

How domain rules become user-facing behavior.

Problem signal	Skill
Bootstrap context	`lamina-business-context`
Inclusive design	`lamina-accessibility`
Copy, labels, errors	`lamina-content-design`
Prioritize, resolve conflicts	`lamina-decision-making`
Affordances, signifiers	`lamina-discoverability`
Domain-empty screens	`lamina-empty-states`
Violation recovery UX	`lamina-error-handling`
Permission/conflict scenarios	`lamina-edge-cases`
Async feedback, status	`lamina-feedback-and-status`
User journeys over state	`lamina-flow-design`
Forms under rules	`lamina-forms`
Entity organization	`lamina-information-architecture`
Wayfinding	`lamina-navigation`
First-run paths	`lamina-onboarding`
UI reflects domain truth	`lamina-product-behavior`
Operations actors perform	`lamina-task-analysis`
Actors, roles, permissions	`lamina-user-modeling`

Agent simulation

Replaces human labs and workshops with contract + live-product simulation.

Problem signal	Skill
Design→build→verify loop	`lamina-design-process`
Scope the design target	`lamina-problem-framing`
Map ask → workflows	`lamina-feature-discovery`
What evidence vs assumption	`lamina-research-scoping`
Plan verify actor walks	`lamina-research-planning`
Merge walk results	`lamina-research-synthesis`
Report verify findings	`lamina-research-communication`
Actor-walk scripts	`lamina-interview-design`
Walkthrough evidence	`lamina-interview-documentation`
Repo + live UI grounding	`lamina-field-research`
Simulated usability (verify)	`lamina-usability-evaluation`
Real metrics only	`lamina-quantitative-validation`
Conflicting actor goals	`lamina-stakeholder-alignment`
User-cited references	`lamina-competitive-analysis`
Workflow priority	`lamina-feature-prioritization`
Group permissions / shared state	`lamina-persuasion-and-groups`
Parallel expert lenses	`lamina-heuristic-review`
Complexity budget / disclosure	`lamina-platform-posture`, `lamina-progressive-disclosure`
Payment / sensitive action honesty	`lamina-trust`
Testable acceptance criteria	`lamina-requirements-definition`

Advanced: Skill files live at skills/lamina-<name>/SKILL.md in the repo. The orchestrator’s audit-profiles.yaml maps workflow sections to specific skills during design and verify.

Direct mode examples


/lamina What invariants should a booking hold enforce?


/lamina How should permission denial appear during checkout?


/lamina What dependencies exist between hotel onboarding and booking?

Next steps

/lamina command — router and direct mode
Orchestrator — how skills are coordinated
Glossary — key terms