Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add emichy/product-judgment --skill prep-customer-interviewgit clone --depth 1 https://github.com/emichy/product-judgmentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/emichy/product-judgment/prep-customer-interview)<a href="https://agentmods.dev/skills/emichy/product-judgment/prep-customer-interview"><img src="https://agentmods.dev/badge/skills/emichy/product-judgment/prep-customer-interview.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00139 | $0.01654 |
| Opus 5 | $0.00069 | $0.00827 |
| Sonnet 5 | $0.00028 | $0.00331 |
| Haiku 4.5 | $0.00014 | $0.00165 |
Grade A, and why
prep-customer-interview scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prep a customer conversation
Working session, not an interview guide. Engineers grab 20 minutes — maybe on a support call, maybe sharing a screen. Every question points at something that already happened — behavior, not hypotheticals.
The trigger is a question, not a launch. Never tell someone to wait until they have something to show.
Customer conversations usually do two jobs: get a reaction to something, and turn up unexpected insights.
House style
Answer first, in one sentence. Then only the evidence that carries it. Evidence is selected, not gathered — the two quotes that carry it, not the six that mention it. Every section earns its place by changing what they'd do this week; if it doesn't, cut it. Slack, not email. No preamble, no applause, and none of "the move" / "the play" / "the tell" / "the key insight here" / "Here's my take."
Output contract
| When | Deliver |
|---|---|
| Always | The anchor, one sentence. Never open with what you need from them, and never as a form or a menu. |
| Call is named | 4–5 questions on the anchor and 1–2 that zoom out — both required. Then Listen for. Skip Who'll react unless they ask who else. |
| Nobody named | Who'll react table and the drafted ask. Skip the script unless they asked for questions too — asked for both, get both. |
| Unclear which | Take the named-call path with a placeholder in the blank. Never ask which they wanted. |
The anchor
One sentence — the thing that, if wrong, changes what you'd build. Not a list of "learning goals."
No bet yet is still a call. The anchor becomes what you don't understand about how they work. Don't manufacture a bet to justify the conversation.
This call is not to count how many people have the problem. Tickets do that. This call is why it hurts and whether they'd change.
The questions
4–5 on the anchor. Then 1–2 that zoom out — not off-topic, wider. Same territory, bigger frame: the job the thing sits inside, how their week actually runs, what they've stopped noticing. The wide ones are not optional. No warm-ups, coaching notes, or hypotheticals as the closer ("if we built X tomorrow…").
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 111 lines · 139 tokens per session scan A e069c5577bcf
prep-customer-interview is a skill published in the GitHub repository emichy/product-judgment (4 stars, last pushed 20d ago), licensed MIT. It adds 139 tokens to every session and 1,654 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
first-customer-finder
Find and qualify evidence-backed potential first customers, early adopters, design partners, or beta users for a startup using recent public pain and buying signals. Use when Codex needs to analyze a product URL or idea, define an ideal customer profile, research public discussions and business pages, identify…
gingiris-user-interview
A playbook for interviewing users and operating an early product launch, including screening participants, running interviews, testing beta versions, and reviewing churn. It is based on a described process for finding product-market fit, meaning a product consistently meeting a real user need.
gr-user-interview
A framework for interviewing users to discover product-market fit, meaning a strong match between a product and a real user need.
run-customer-discovery
Use when validating a startup idea, product hypothesis, or new feature by conducting structured interviews with target customers before building.
jtbd-extractor
Turn raw research into Jobs-to-be-Done statements showing what users are really trying to accomplish. Use when: extract jobs, jtbd analysis, jobs to be done, what job is the user hiring, underlying user needs.
first-customer-finder
Find and qualify evidence-backed potential first customers, early adopters, design partners, or beta users for a startup from recent public signals. Trigger on "find my first customers", "who would buy this", "find early adopters", "find design partners", "find beta users", "find leads for my startup", or when given a…