Skip to content
Agent Orchestration
Skill

/lamina-product-behavior

Define authoritative runtime product truth across actors and views. Use when modeling domain behavior, setting product density, specifying side effects, protecting invariants, choosing consistency guarantees, handling retries or concurrent actions, defining time semantics and

BOOST
From plugin
lamina
12010 skills
Install
$ npx -y skills add aryaniyaps/lamina --skill lamina-product-behavior --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/lamina-product-behavior

Context preview

The summary Claude sees to decide when to auto-load this skill.

Define authoritative runtime product truth across actors and views. Use when modeling domain behavior, setting product density, specifying side effects, protecting invariants, choosing consistency guarantees, handling retries or concurrent actions, defining time semantics and

SKILL.md

lamina-product-behavior.SKILL.md
name: lamina-product-behavior
description: "Define authoritative runtime product truth across actors and views. Use when modeling domain behavior, setting product density, specifying side effects, protecting invariants, choosing consistency guarantees, handling retries or concurrent actions, defining time semantics and prerequisites, drawing ownership boundaries, or keeping views coherent. Use lamina-systems for causal loops and change over time, and lamina-ux for interaction expression."

Lamina Product Behavior

Reference-loading protocol

1. Match the request's primary runtime-truth decision to one row below. 2. Open that linked reference before answering. Add another only when a second decision materially changes the answer; do not preload the directory. 3. Start the response with `Using lamina-product-behavior: <topic path(s)>` so the selected behavior lens is auditable.

Topic index

| Runtime signal | Read | Adds | |---|---|---| | Need the represented domain model to match how actors understand the task | [Product Behavior](references/product-behavior.md) | entities, permissions, workflows, surfaces, and illegal states | | Need to choose how much power or complexity each actor sees | [Product Density](references/platform-posture.md) | actor- and surface-specific complexity budgets | | One mutation must update, notify, invalidate, audit, or retry elsewhere | [Side Effects](references/side-effects.md) | primary effect, downstream effects, delivery lifecycle, and failure recovery | | Need a rule that must hold in every valid state | [Invariants](references/invariants.md) | explicit predicates and violation probes | | Views may lag or disagree after a state change | [Consistency Guarantees](references/consistency-guarantees.md) | strong, eventual, and read-your-writes product promises | | An operation may repeat, race, conflict, or arrive after state changed | [Idempotency and Concurrency](references/idempotency-concurrency.md) | duplicate safety, conflict policy, and fencing | | Behavior depends on zones, dates, recurrence, duration, deadline, or correction | [Time Semantics](references/time-semantics.md) | explicit temporal meaning and boundary behavior | | A workflow requires prior setup, ownership, state, or another workflow | [Dependencies](references/dependencies.md) | prerequisite edges and unmet behavior | | Need to decide which domain owns a rule or what data crosses a boundary | [Modularity Boundaries](references/modularity-boundaries.md) | information hiding and cohesive ownership | | Several roles or surfaces must agree on the same entity and lifecycle | [Multi-View Integrity](references/multi-view-integrity.md) | handoffs, visibility, and cross-view truth |

Working rule

Use the smallest sufficient reference set. Common pairs are invariants + idempotency for consequential writes, consistency + multi-view integrity for role-based products, and side effects + time semantics for scheduled delivery. Keep behavior in product language; do not prescribe implementation machinery unless the user asks for it.

Read more
Ships withlamina

Headless product design for AI coding agents, backed by a transactional product graph | Design how it works, verify what you ship.

Get the whole plugin

Other skills on lamina.