# 32 Claude Skills for Product Interviews, From the Job Ad to the Offer 32 Claude skills to read the role, turn your real work into stories, practise each interview round, answer how you use AI, and compare offers against your own criteria. - Page: https://lantern.meet-polar-bear.com/skills/product-interviews-pack - Download: https://lantern.meet-polar-bear.com/skills/product-interviews-pack.zip - Source: https://github.com/polar-bear-org/claude-skills/tree/main/packs/product-interviews-pack - Price: free ## Install in Claude Code ``` /plugin marketplace add polar-bear-org/claude-skills /plugin install product-interviews-pack@polar-bear-skills ``` ## The 32 skills you get The set follows one job search in eight stages: reading the role, telling your story, preparing the cases, preparing the conversations, showing how you work with AI, getting in through people, running the process, then deciding. ### 1 · Read the role - **Target Role Brief** — use when: You are applying to everything and cannot say in one line what you want next. Output: Two or three target roles with level, your criteria and deal-breakers, the words real ads use, each matched to your evidence. - **Job Ad Breakdown** — use when: The ad lists twenty requirements and you cannot tell which three decide the hire. Output: One ad read line by line: the problem behind the role, must-haves, level signals, the AI ask, implied rounds, your evidence, gaps. - **Company Research Brief** — use when: The interview is in three days and you could only describe the logo. Output: What the product does and for whom, your own notes from using it, dated news with sources, three questions a website cannot answer. - **Interviewer Research Notes** — use when: You have five names on a calendar invite and know a job title for each. Output: Per interviewer: public professional facts with source and date, likely angle by role, unknowns, one question only they could answer. - **Interview Loop Map** — use when: Six rounds are booked and you are preparing the same answers for all of them. Output: Every round with date, length, interviewer, format and what it assesses, the AI rule per stage, stories assigned, questions for the recruiter. ### 2 · Tell your story - **Career Evidence Inventory** — use when: Fifteen years of work and you cannot remember what you did in the role before last. Output: One evidence row per piece of work, the unglamorous work logged on purpose, strengths backed by two rows, a do-not-claim list. - **Interview Story Bank** — use when: Every behavioural question gets the same launch story, told a little differently each time. Output: Eight to twelve stories in a two minute and a thirty second version, tagged by competency, with follow-ups and facts not to stretch. - **Tell Me About Yourself** — use when: The first question of every interview is the one you answer worst. Output: A ninety second answer in present, past, future order, a twenty second version, two true sentences for "why are you looking". - **Product Resume Bullets** — use when: Your resume is a list of launches and reads like the job description. Output: Two to four lines per role as challenge, action, result, with scope, your decision, your part apart from the team's, sourced numbers only. - **LinkedIn Profile Rewrite** — use when: Your profile undersells senior work, or a chatbot rewrote it and it reads like everyone else's. Output: A section check, three headline options, an About in your own voice, a claim map, mismatches with the resume to fix first. ### 3 · Prepare the cases - **Product Sense Case** — use when: You know the steps by heart and your answers still sound like a recited framework. Output: A practice case on an open product or design question, then notes on your answer against a rubric you can see. - **Metrics Case** — use when: You say "I would look at the data" and cannot say which data or what would change your mind. Output: Practice on choosing success metrics and on diagnosing a metric that moved, with your reasoning written out and checked. - **Prioritisation and Execution Case** — use when: The question is "you have five things and room for two" and you answer by listing all five. Output: Practice on a trade-off question: options, what each costs and who pays, the call, the cut, the risk you accept. - **App Critique** — use when: You practise by criticising random apps and it never turns into a method. Output: A practice critique of an app you choose: users, how it earns its place, flows walked through, the one change you would argue for. ### 4 · Prepare the conversations - **Behavioural Interview Answers** — use when: A values or leadership round is next and your stories all say "we". Output: For one round: the competencies tested, a story and a reserve story for each, the follow-ups and your honest answers. - **Portfolio and Project Walkthrough** — use when: You stopped designing every screen or writing every spec, and your walkthrough still shows screens and timelines. Output: A timed talk on two or three pieces of work: stakes, your role, three decisions, the mistake, the outcome, a slide outline, likely questions. - **Questions to Ask** — use when: "Do you have any questions for us" and you ask about the culture. Output: Questions for each interviewer by role, website-answerable ones removed, three marked to ask first. - **Mock Interview** — use when: You have prepared on paper and never said any of it out loud. Output: A practice round with a stand-in interviewer and follow-ups, then notes on each answer as present, thin or missing. ### 5 · Show how you work with AI - **Candidate AI Use Plan** — use when: You do not know whether using AI on the take-home is expected or forbidden, and you have not asked. Output: One page per employer: what they allow at each stage and where it is written, the question for the recruiter, the sentence you will say. - **AI Fluency Answer** — use when: The ad asks for AI fluency, your title has never had AI in it, and your answer is a list of tools. Output: An answer to "how do you use AI in your work" from two or three real examples, in a short and a long version, with no tool list. - **AI Work Sample** — use when: They will ask you to show it, not describe it. Output: A one page write-up of a real task, an optional small prototype labelled as a prototype, a confidentiality check. ### 6 · Get in through people - **Target Company List** — use when: Your list is five famous names and you have never checked who you know at any of them. Output: Ten to twenty companies with a sourced reason each fits, the open role, who you know there, and a dropped list. - **Reason to Write** — use when: You have not spoken in six years and every draft turns into an ask by line two. Output: For each former colleague, one honest reason to get back in touch, what not to ask yet, points for a message you write. - **Informational Interview Questions** — use when: An insider agreed to twenty minutes and you are about to ask what the careers page already says. Output: Questions for a conversation with an insider, website-answerable ones removed, five to ask first, a closing question. - **Referral Ask** — use when: A role opened where a former colleague works and you are not sure you can ask. Output: A readiness check on the relationship and the role, a plain "not yet" when it is, points for asking with an easy no. ### 7 · Run the process - **Recruiter Screen Prep** — use when: A recruiter wants thirty minutes tomorrow and will ask what salary you expect. Output: A check that an inbound message is real, then one page for the first call: why you are looking, timing, pay expectations, your questions. - **Take-Home Assignment Plan** — use when: The take-home says "spend as long as you need" and you are about to give it a week. Output: The brief restated, questions to send first, a time box you set, the submission structure, an assumptions page, a note on AI use. - **Interview Round Debrief** — use when: You walked out, wrote nothing down, and gave the same weak answer in the next loop. Output: What was asked, what you answered, promises made, what to change in your stories, points for a specific thank-you. - **Job Search Tracker** — use when: Five processes at five stages and you just missed a follow-up. Output: One row per process, a twenty minute weekly review, one nudge for a silent process and then a close. ### 8 · Decide - **Offer Comparison** — use when: Two offers, a deadline on one, and you are comparing base pay only. Output: Each offer, and staying where you are, against your own criteria and weights, with sources, unknowns and questions before you sign. - **Offer Negotiation Prep** — use when: The offer is good and you know you should ask, and you have nothing but a feeling. Output: Your real alternative, what you want and the least you would accept, the facts behind each ask, the order of the conversation. - **30-60-90 Day Plan** — use when: You start in three weeks, or a final round asks what you would do in your first ninety days. Output: A first draft for the new role: who to meet, what to learn before changing anything, one early result, what you will propose. ## Workflows - **Read one role** (An ad you like and an interview that may follow.) Target Role Brief → Job Ad Breakdown → Company Research Brief → Interviewer Research Notes → Interview Loop Map - **Turn work into stories** (Years of shipped work and a resume that lists launches.) Career Evidence Inventory → Interview Story Bank → Tell Me About Yourself → Product Resume Bullets → LinkedIn Profile Rewrite - **Practise the cases** (A case round is booked and your answers sound recited.) Product Sense Case → Metrics Case → Prioritisation and Execution Case → App Critique - **Prepare the conversations** (A values round or a panel walkthrough is next.) Behavioural Interview Answers → Portfolio and Project Walkthrough → Questions to Ask → Mock Interview - **Show your AI work** (The ad asks for AI fluency and the rules on AI use are unclear.) Candidate AI Use Plan → AI Fluency Answer → AI Work Sample - **Get in through people** (A list of five famous names and former colleagues you have not written to in years.) Target Company List → Reason to Write → Informational Interview Questions → Referral Ask → Recruiter Screen Prep - **Run it to a decision** (Several processes at different stages, and an offer on the way.) Take-Home Assignment Plan → Interview Round Debrief → Job Search Tracker → Offer Comparison → Offer Negotiation Prep → 30-60-90 Day Plan ## The quality bar - Sourced claims: every claim about you traces to an evidence row you can source. A missing number stays a placeholder - Your part, not the team's: your part is kept apart from the team's. Colleagues appear by role and nobody is judged - No scoring of you: nothing scores, ranks or grades you. Practice notes rate the answer against a rubric you can see, as present, thin or missing. Offers are scored against your own criteria, and you decide - Public facts only: facts about companies and people come from public pages, with a source and a date. Unknown stays unknown. Nothing about an interviewer is kept in memory - Points, not messages: messages to people are given as points. You write and send them yourself - No pay figure from Claude: figures come from the employer's stated range or from you, with a source and a date, and legal points on offers, contracts, recording and personal data end with “check with a qualified adviser” - The red line: Claude helps you prepare and practise from work you really did: it never invents a role, a result or a number, it stays out of the live interview, and when an employer asks what you used AI for, you tell them ## Questions people ask ### What is in the Claude Guide for Product Interviews pack? 32 Claude skills that follow one job search in eight stages: read the role, tell your story, prepare the cases, prepare the conversations, show how you work with AI, get in through people, run the process, and decide. Each skill is one method or one artifact. ### How do I install these Claude skills? In Claude Code, two lines install all 32 at once. In Claude.ai or Claude Desktop, download the zip, open Settings, Skills, Add, and upload the skills you want. Either way it takes about five minutes. ### Do I need Claude Code to use them? No. The skills work in Claude.ai, Claude Desktop and Claude Code. In the app you upload them yourself; in Claude Code one command installs the whole pack. ### Do the skills work with ChatGPT or other assistants? They are built as Claude Skills, and installing them is a Claude feature. Each one is a plain Markdown file, so you can read it and reuse the method elsewhere, but it will not load and trigger by itself outside Claude. ### Can Claude help me during the live interview? No. Claude helps you prepare and practise from work you really did. It never invents a role, a result or a number, it stays out of the live interview, and when an employer asks what you used AI for, you tell them. ### Is the pack free? Yes. It is free to use inside your company and in client work, and not for resale. There is no signup: the download and the install command are on this page. ## About Written by Polar Bear (https://meet-polar-bear.com/). Published on Lantern (https://lantern.meet-polar-bear.com/), which turns your notes, documents and call transcripts into social post drafts you review and publish.