メニュー
Structure a spoken PM product-sense answer with assumptions, segmentation, pain-point prioritization, and MVP tradeoffs. Use when practicing design, improve, or build-next interview questions.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/TuYv/ccpm --skill product-sense-interview-answerコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Use when auditing a paid ad account for incremental contribution, wasted spend, or measurement integrity before scaling; runs a typed 20-item ROAS profile with verified vetoes and a SHIP/FIX/BLOCK/UNDECIDED gate on own exported data. Not for campaign structure design — use campaign-architect; not for creative production — use ad-creative-builder. 付费广告账户审计/ROAS评分
Use when the user asks to "write ad copy", "generate RSA headlines", or "build ad creative at volume"; produces ad units — RSA headlines/descriptions, hooks, and an angle matrix — message-matched to the destination landing page. Not for scoring an ad account — use ad-account-auditor; not for the post-click page — use landing-optimizer; not for organic articles — use content-writer. 广告创意/广告文案/RSA标题
Use when the user asks to "design an A/B test", "set up a creative/landing test", "run an incrementality test", or "is this result statistically and practically material?"; produces a hypothesis, variant matrix, sample-size/duration/power plan, and a documented effect/uncertainty read from own exported results. It applies only a precommitted owner-approved action rule; the statistical helper never chooses a business action. Not for producing variants — use ad-creative-builder; not for reading back one shipped change — use paid-measurement-loop. 广告AB测试设计/实验设计/显著性判定/增效测试
SOC 職業分類に基づく
SKILL.md を表示中
| name | product-sense-interview-answer |
| argument-hint | [interview prompt] |
| description | Structure a spoken PM product-sense answer with assumptions, segmentation, pain-point prioritization, and MVP tradeoffs. Use when practicing design, improve, or build-next interview questions. |
| intent | Coach PM candidates through open-ended product-sense interviews using a repeatable six-part answer spine: clarify, rationale, goal, segmentation, pain points, and solution choice. Use this to practice product design and product improvement questions, avoid solution-first answers, and produce responses that sound thoughtful out loud rather than over-scripted on the page. |
| type | component |
| theme | career-leadership |
| best_for | ["Practicing product design and product improvement interview questions","Coaching candidates who jump to solutions too quickly","Turning messy ideation into a crisp spoken interview answer"] |
| scenarios | ["How would you improve YouTube?","Design a product for travelers with flight anxiety","What would you build next for DoorDash?"] |
| estimated_time | 20-30 min |
Help PM candidates and interview coaches structure product-sense answers that sound strong out loud, not just on paper. Use this when practicing prompts like "How would you improve X?", "Design a product for Y", or "What would you build next for Z?"
This is not a memorize-and-recite script. It is a reasoning scaffold that prevents solution-jumping, forces real prioritization, and leaves the interviewer with a clean story they can follow.
Works best with: The interview prompt you're practicing (e.g., 'How would you improve X?', 'Design a product for Y'). Also useful: The company/role you're interviewing for and how much time the answer gets.
Anything supplied with the invocation itself — text after the skill name, a pasted context dump, or an appended ARGUMENTS: line — counts as answers already given. Use it and skip whatever it covers; don't re-ask.
Arriving empty-handed? That works too. The skill asks for the prompt, then walks the reasoning scaffold with you out loud.
Example invocation: Practice this: 'How would you improve Google Maps for commuters?' — 25-minute answer, L5 PM loop.
Strong product-sense answers do more than generate ideas. Interviewers are usually testing whether you can:
The order matters. If you skip from prompt to feature ideas, your answer sounds clever but ungrounded. If you establish the user, goal, and pain first, your solution feels earned.
Use template.md as the working structure.
Start by surfacing the two ambiguities that change the answer most. Good clarifiers usually narrow:
If the interviewer does not answer, state your assumptions and move on. The goal is to unblock the rest of the answer, not to turn the interview into requirements gathering.
Quality bar: Ask questions that materially change the solution. "Are we talking mobile or desktop?" matters less than "Are we optimizing for viewers, creators, or advertisers?"
Explain why the space matters now.
For the market view, cover:
If a company is named, then add:
End this section with a one-line thesis. That thesis should make the rest of the answer feel inevitable.
Quality bar: Use qualitative signals unless you know the numbers cold. Fake precision is worse than grounded judgment.
Write one sentence in this format:
Help [user] [achieve outcome], so that [broader impact].
Then describe what success looks like for the user in observable terms.
Good: "Help beginner YouTube learners find content they are glad they watched, so that the platform becomes an intentional learning destination."
Bad: "Build a personalized AI learning path feature." That is a solution disguised as a goal.
Do not jump straight to persona. First identify the ecosystem players, then choose the player you want to serve. After that, choose two segmentation dimensions that actually change needs.
Good segmentation dimensions usually change:
Weak dimensions are often demographic cuts that do not change the product meaningfully.
After choosing your target segment:
Break the user journey into 4-6 stages. Then list the frictions across that journey.
Prioritize the top pain point using:
This is the fulcrum of the entire answer. If the pain point is vague or weak, the solution section becomes generic.
Quality bar: Pain points should describe user friction, not missing features. "No structured progression after each video" is a pain. "No AI learning path" is already a solution.
List three distinct solutions. They should solve the same pain in different ways, not represent three feature line-items inside one idea.
Evaluate each option on:
Then choose one MVP and specify:
Close with a one-sentence recap that names:
That final sentence is what the interviewer should remember.
See examples/improve-youtube.md for a full worked example.
What makes it strong:
A strong answer to this prompt would explicitly state a startup assumption if no company is named, prioritize people who live alone, and choose the wake-up problem before discussing dispatch or smart-home integrations.
What makes this example useful:
"I would improve YouTube by adding AI summaries, better recommendations, creator analytics, and a study mode."
Why this fails:
This kind of answer can sound energetic in the moment, but it signals weak PM judgment.
template.mdexamples/improve-youtube.mdskills/problem-statement/SKILL.mdskills/proto-persona/SKILL.mdskills/customer-journey-map/SKILL.mdskills/opportunity-solution-tree/SKILL.md