jamiemill/layers-skills
9 skills5.2k installs1.4k starsGitHub
Install
npx skills add https://github.com/jamiemill/layers-skillsSkills in this repo
1Layers Conceptual ModelLayers Conceptual Model is an agent skill for solo builders and small teams who need a load-bearing product layer that most specs skip: which entities exist, how they relate, what states they can enter, and the vocabulary everyone will use. It treats the model as a design decision in solution space—not a dump of contradictory user mental models and not an engineer-owned schema diagram. The skill guides questions about object boundaries, relationships, and lifecycle states without prescribing UI flows or navigation. That separation helps indie founders align Claude or Cursor agents on one coherent story before frontend, backend, and docs work diverge. Because the README stresses that unexamined gaps between this model and what ships become both UX debt and evolution pain, it fits early validate scope work and stays relevant through build PM when you reconcile specs with reality. Pair it after layers-intro when you are structuring a new SaaS, agent product, or app and want agents to reason about the same nouns and rules you will ship.586installs2Layers Interaction Flowlayers-interaction-flow is a Design-layer skill for solo builders and small teams who want honest interaction maps before visual polish or implementation tickets. It defines how someone moves through places, what they can do, what content each state needs, and what happens when things fail—without choosing typography or color. The skill is explicitly a technique library paired with layers-intro, so agents pick methods rather than running a single checklist. It warns that unstable conceptual models underneath will waste rework, and it encodes product-design discipline: every action names a destination, and edge cases are first-class. Use it when a feature feels “designed” in mockups but handoffs still argue about states, or when you need a shareable flow artifact for Validate scope reviews and early Build PM alignment.581installs3Layers Product Strategylayers-product-strategy is a Layers methodology skill for solo builders who need product direction without enterprise PM overhead. It sits at the opening of the solution space: translate what you learned about users into deliberate choices about which outcomes matter, which pains deserve investment, and which solution bets you will test first. Unlike a feature checklist generator, it supplies disciplines—measurable outcomes, opportunity statements tied to journey moments, and cheap experiments on the riskiest assumptions—while reminding you to stop when decisions are already obvious. Load it after layers-intro when you are deciding what to build next, reprioritizing a roadmap, or aligning a thin MVP with a single north-star metric. The skill expects judgment: pick techniques that fit your context rather than running every exercise. Outputs are decision artifacts (outcome, opportunity map, bet list, test plan) you can hand to writing-plans, prototyping, or validation landing work in later journey phases.580installs4Layers Domainlayers-domain is a Layers methodology skill that teaches agents how to map a domain’s concepts, language clashes, and natural bounded contexts before design commitments. Solo builders shipping complex B2B or workflow products use it when user interviews, docs, and stakeholder jargon disagree, and they need a disciplined observation layer instead of smoothing conflicts too early. The skill frames domain work as real-world mapping: capture synonyms and polysemy, document where communities diverge, and note event-shaped processes—without resolving those tensions here. When the domain is already stable, the readme allows skipping straight to conceptual modeling. It is a technique library, not a fixed script, and expects familiarity with the broader Layers intro sequence.579installs5Layers OrientLayers Orient is a journey-wide diagnostic skill from the Layers framework for solo builders and designers who feel stuck but cannot name why. In one light session it maps whether key decisions exist at each of seven layers, flags decisions that look solid but are still assumptions, and names the bottleneck—the most foundational layer still unresolved, which puts everything above it at risk. It deliberately avoids becoming a long report; success means a specifically named layer, honest strength versus assumption labels, and a pointed handoff to the right layer skill such as strategy, interaction, or visual work. Load layers-intro first for shared vocabulary, then say "let's orient" or describe what prompted you to pick up the design. Use orient whenever priorities feel equal across UX, brand, and implementation fights, not only at project kickoff. It is methodology for sequencing design work, not a wireframing or codegen tool.579installs6Layers Surfacelayers-surface is a Design & UI/UX agent skill from the Layers stack for solo builders and small teams shipping SaaS, mobile, or multi-channel products. It assumes layers-intro is already loaded and supplies a library of audit techniques for the surface layer: words, motion, feedback, prominence, and accessibility across screen, voice, or physical touchpoints. Use it when polish feels wrong but you are unsure whether to tweak copy and layout or reopen strategy and structure underneath. The skill trains the central judgment called out in the doc—surface symptoms versus deeper-layer gaps—so you do not paper over missing objects or broken jobs-to-be-done with prettier UI. It is most often invoked during frontend implementation and pre-ship review, but the same checklist helps validate prototypes and launch-ready flows when you need consistency before users see the product.579installs7Layers Introlayers-intro is the mandatory orientation skill for Jamie Mill’s Layers of Product Design family. Solo and indie builders invoke it at the beginning of a design session so Claude Code, Cursor, Codex, or similar agents share the same vocabulary: seven layers from reality through problem knowledge (behaviour, domain, needs) into deliberate solution choices (strategy, conceptual model, interaction flow, surface). The framework stresses that layers are not a waterfall—you can start anywhere—but weak lower layers create UX debt that shows up in UI and copy later. That makes the skill journey-wide: useful when validating scope, planning features, structuring flows, or polishing surfaces, as long as you have not yet aligned on the shared model. It does not produce pixels or code; it supplies the context graph every other layers skill assumes, which reduces contradictory design advice mid-session and keeps agent output tied to user needs rather than ad-hoc UI opinions.576installs8Layers Observed BehaviourLayers Observed Behaviour is a journey-wide agent skill from the Layers research methodology for planning studies and synthesizing evidence into what users actually do—not what the team wishes they did. Solo builders use it whenever discovery, validation, or iteration needs honest grounding: at idea time to design interviews and surveys, during validate to turn partial notes into patterns, and later in grow or operate when support logs or usage data need the same confidence discipline. The skill forces a fork between Plan (no research yet) and Synthesise (material exists), with guidance to synthesize first then plan gap-filling when research is incomplete. Highlights include staying near verbatims, marking confidence tiers, and deciding which questions matter most. It is editorial and process-oriented rather than a code integration; load layers-intro before this library skill. Best for product-minded indies who want research artifacts agents can extend into specs, landing copy, or backlog items without collapsing observation into opinion.573installs9Layers User NeedsLayers User Needs is a journey-wide facilitation skill for solo builders and small teams practicing the Layers product framework. It treats user needs as interpreted opportunities—not raw telemetry—and elicits needs, pains, and desires using job stories by default so situations stay concrete and solution bias stays out. The skill forces decisions about which users matter, what jobs they hire the product for across functional, emotional, and social dimensions, what is evidence-backed versus assumed, and what to prioritize next for strategy. When you already run Agile ceremonies, you can swap in user stories; when you have a large installed base, top tasks analysis offers a statistical complement. Outputs align with opportunity solution tree thinking so downstream layers can map solutions deliberately. Load layers-intro first for shared vocabulary. Reach for this skill whenever you are about to commit roadmap or spec work but only have fuzzy personas or feature wishlists—early ideation, validation pivots, pre-build PRDs, or post-launch iteration reviews all benefit from the same structured elicitation.571installs