define-product
Define what a product is and why it exists, by orchestrating Matt Pocock's grilling and domain-modeling skills (plus research, competitive briefs, logic prototypes and stakeholder questionnaires when needed), and record the result in an impeccable-compatible PRODUCT.md and CONTEXT.md. Product phase called by start-project or askmatt, or when the user explicitly asks to define the product.
- License 许可
- MIT
- 最近更新
- 2026-10-07
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- define-product
- description
- Define what a product is and why it exists, by orchestrating Matt Pocock's grilling and domain-modeling skills (plus research, competitive briefs, logic prototypes and stakeholder questionnaires when needed), and record the result in an impeccable-compatible PRODUCT.md and CONTEXT.md. Product phase called by start-project or askmatt, or when the user explicitly asks to define the product.
02 资源文件 1
- 参考文档references/question-bank.md2.5 KB
03 SKILL.md 正文
Define Product
Orchestrate; don't improvise the method. The interview engine is product-studio:grilling, the language discipline is product-studio:domain-modeling. This skill adds the no-coder framing, the order, and the outputs. Danish in conversation, English in files.
Order of play
- Load the engines. Invoke
product-studio:grillingandproduct-studio:domain-modeling(Matt'sgrill-with-docspairing). If refused as user-only, read their SKILL.md files and follow them. From here on their rules govern: design tree, frontier rounds, numbered questions each with➡️a recommended answer, facts are yours and decisions are the user's, CONTEXT.md updated the moment a term settles, ADRs only when hard to reverse + surprising + a real tradeoff. - Write back the idea (brainstorming principle): 3-6 bullets on outcome, audience, success; "what you said" vs "what I assume". Correct before going on. This is the tree's root.
- Raw idea? If the user can't yet say who it's for or what problem it solves, first run
product-studio:idea-refine(user-only: read../idea-refine/SKILL.md): variations (inversion, 10x simpler, another audience), painkiller vs vitamin, and a Not Doing list. Its output seeds the tree; it is not saved todocs/ideas/.
- Raw idea? If the user can't yet say who it's for or what problem it solves, first run
- Scope check. Several independent products hiding in one idea? Say so, pick the first. Too foggy for one session? Hand to
product-studio:wayfinder(the start-project conductor resumes after). - Grill. Seed the design tree from
references/question-bank.md. Ask the full frontier each round; cap at ~7 questions, holding the rest for the next round. Plain Danish, concrete scenarios ("Mette fra økonomi logger ind mandag og ser 40 ubetalte fakturaer…"). Challenge fuzzy words on the spot ("kunde", "bruger", "ordre", "konto"). Three extra habits: add a confidence line to your recommendation when you're guessing ("➡️ B, ~60% sikker, fordi…"); ask once per product "hvis du ikke skulle forsvare det over for nogen, hvad ville du så bygge?" to separate want from should-want; and treat "lyder godt", "hvad du synes" or silence as not yet a decision: restate and ask for the explicit choice. - Pull in specialists only when the tree needs them (run them without blocking the rest of the round):
- Outside facts (regulation, competitors' pricing, an API's limits) →
product-studio:researchin the background. - Competitive landscape matters to positioning →
product-management:competitive-briefif installed; otherwiseresearch. - A business rule can't be settled in words (states, approvals, calculations) →
product-studio:prototype(logic branch): a single clickable HTML file the user can drive, published as an artifact. - The answer sits with someone else (a customer, an accountant) →
product-studio:to-questionnaire.
- Outside facts (regulation, competitors' pricing, an API's limits) →
- MVP cuts. When jobs and users are clear, offer 2-3 MVP cuts (smallest useful / sellable v1 / demo that wins the meeting) with your recommendation. YAGNI.
- Write
PRODUCT.md(with an explicit Out of scope / Non-goals list, including idea-refine's Not Doing items) in the impeccable schema from../start-project/references/docs-layout.md(root of the repo; update rather than replace an existing one). Keep it readable for a non-developer; no technology. - Gate. ≤10-bullet summary + file list: "Er det her produktet? Skriv godkendt, eller ret mig." The frontier must be empty: every branch visited, nothing silently assumed.
Done when
PRODUCT.md and CONTEXT.md reflect the approved understanding, open questions name where they'll be resolved, and the user said "godkendt".
04 同仓库更多 Skills
writing-skills
Use when creating new skills, editing existing skills, or verifying skills work before deployment [Superpowers; user-invoked. Use to write or pressure-test a skill, e.g. product-studio changes.]
writing-plans
Use when you have a spec or requirements for a multi-step task, before touching code [product-studio: canonical Code-tab execution, entered via build-ticket; user-invoked otherwise]
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
wizard
Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.
webapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs. [Python alternative to playwright-skill (canonical for scripted browser checks); user-invoked only.]
web-design-guidelines
Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".