New here, or not sure this is the right skill? Start right here — or run

/nmt-chat

, describe your situation, and it points you to the right one. Quick map: new idea →

nmt-market-research

· live product or a metric moved →

nmt-diagnose

· have customer interviews →

nmt-analyze-interviews

· ready to build →

nmt-product-requirements

· positioning / launch copy →

nmt-craft-value-proposition

→

nmt-craft-go-to-market

.

• Is: a conversational diagnostic. Through ≤15 adaptive questions it challenges your goal, then surfaces all risks, all growth points, and the risky assumptions in your current initiatives, prioritizes the first move, and points you to the skill that does it.
• Is not: a report generator. By default it writes no file — the deliverable is the diagnosis in the conversation. It writes one file only if you ask.
• Is not: an auto-launcher. It recommends the next skill and says why; you launch it.
• Is not: a generic feature-by-feature audit. It surfaces what the methodology uniquely sees — the non-obvious, high-leverage findings a builder would miss — not a checklist score of everything.

CLAUDE.md

• A Job ≠ "progress." A Job specifies a desired transition — situation (State A) → expected outcome (State B), in order to perform a higher-level Job. A unit of motivation.
• Value = greater energy efficiency for the brain in performing a Job, vs. the brain's prediction. The Aha Moment is the customer-experience of value beating prediction; the Problem is value below it. Never use the abbreviations PPE/NPE (Rule 22).

• I want to + verb

  is the primary element, not the whole Job (eight elements). Each infinitive verb is a separate Job (Rule 7).
• A Problem ≠ a root cause — it's the consequence of a Solution hired for a Job and underperforming its success criteria.
• A Solution is a thing in the world AND a label for the sub-graph it installs.

the-algorithm.md §2

../nmt-chat/references/Next-Move-Theory-Canon/Advanced-Jobs-To-Be-Done/ajtbd-key-theses.md

../nmt-chat/references/Next-Move-Theory-Canon/Riskiest-Assumption-Test/rat-key-theses.md

../nmt-chat/references/Next-Move-Theory-Canon/Algorithms/the-algorithm.md

../nmt-chat/references/Next-Move-Theory-Canon/Next-Move-Theory/nmt-key-theses.md

ABCDX-Segmentation/abcdx-segmentation-key-theses.md

Advanced-Jobs-To-Be-Done/segmentation.md

Advanced-Jobs-To-Be-Done/value-creation-mechanics.md

Advanced-Jobs-To-Be-Done/job-graph.md

Advanced-Jobs-To-Be-Done/value-creation.md

Advanced-Jobs-To-Be-Done/behaviour-change.md

Path note. Use

../nmt-chat/references/Next-Move-Theory-Canon/...

; if not found, retry with a

1-

prefix (

1-Next-Move-Theory-Canon/...

) — the source repo numbers folders, the public mirror strips the prefix.

What this does: finds the risks and growth moves holding your product back — including risky guesses in what you're already doing — then points you to the right next step. A few quick questions (more only if it helps). You decide: you confirm the goal and launch the recommended skill; I don't run it for you. Honest caveat: every finding is a guess to check — the cheapest check comes first.

• Ask "Why do you want this? In order to do what?" and trace it up 3–5 levels (5 Whys) — from feature/metric to the bigger business result it really serves: conversion, sales, margin, profit, the strategic goal (in the method, climbing their business-Job graph).
• At each level up, look for a better move — a more interesting or lower-effort way to hit that bigger result than the task they came in with. The real next move often sits one or two levels above the stated task.
• Tune vs. bigger bet: open to changing segment / business model / market (a bigger bet — usually a founder / C-level call), or improving the current product (tuning what you have)? The two run in parallel, not as opposites.

• Stage / PMF: idea (PMF 0) · early, few payers · paying base, weak PMF · strong PMF, scaling. The master branch — PMF 0 routes mostly to

  /nmt-market-research

  ; a live product opens the ABCDX path.
• Product in one line + the Core Job hypothesis (what people hire it for).
• B2C / B2B.
• Current initiatives / roadmap: "What are you doing or planning right now about this — features, bets, experiments?" Capture everything; each becomes a target for the RAT pass (every initiative is a stack of assumptions — extract and flag the riskiest). Skip if none.

• Market: real money in this Job, or a tiny segment? (size · frequency)
• Segment+Job: do you know your A-segment by their Jobs (ABCDX run)? who pays and why? "why do they stay?"
• Value: what do you do noticeably better than the current way — do customers feel it (the Aha)? satisfaction level?
• Unit economics: does one customer's margin work (price − cost to serve)?
• Demand & acquisition: reachable at acceptable CAC? do they understand the value (activating knowledge)?
• Activation: do new users reach the Aha Moment — what % and how fast?
• Retention / repeat: churn, repeat rate, where the Critical Chain of Jobs breaks.

1. The goal, checked — the task confirmed, or a reframed higher-leverage goal with the climb that justifies it ("you asked how to lift X; the bigger result it really serves is Y, and that's the better thing to move").
2. Tackle this first — the single highest-leverage move and why, in one or two sentences.
3. The cheapest way to check it — every finding is a hypothesis; the one cheap test to run before committing. Then the route: the skill that does this first move + the exact next action (you launch it). See the routing table.

1. Top risks — the top 3 weak spots, each traced to its real upstream cause (not the symptom), binding one first. Any others: one line each.
2. Top growth moves — the top 3 ways to grow, plain-first, one line each; any others as a single line. The moves, in plain terms (term in parentheses):
   • remove a step they hate — make an unwanted task disappear for the whole group, for good (kill a Job)
   • grab what they do right before or after you — own the must-do task just ahead of or just behind the one you already do (Previous / Next Job)
   • do the bigger task for them — make the bigger task the thing your product fully does, so a layer of small steps vanishes (move up a level)
   • serve a nearby group you're ignoring — a sibling task right beside what you do, for a group next door (adjacent Small Jobs)
   • nail a need everyone handles badly — a concrete bar the whole market underserves today (underserved success criteria)
3. Risky assumptions in what you're already doing — one row per current initiative: what you're doing · what it quietly assumes · the assumption that sinks it if wrong · cheapest way to check. (Omit if none described.)

../nmt-chat/references/producer-contract.md

• §1 Helicopter-view — yes (flow step 0).
• §3 Input-as-hypothesis — yes: everything the user reports (metrics, segment beliefs, initiatives) is a claim, not a fact; the diagnosis flags where a "fact" is actually unmeasured, and the current-initiative RAT pass is exactly this gate applied to their plans.
• §4 Validation framing — yes as the per-finding "cheapest validation step"; no separate validation-debt counter unless a file is written.
• §2 output format / §5 output path — only when the user asks to save (then

  .md

  /

  .html

  + custom path apply).
• §6 Deep-mode QA / web-MCP — N/A by default (the diagnosis is reasoning over the user's data; web research is the routed skill's job).

• Language. Default English; if the user writes in another language, offer to continue in it, then hold it. Canon files and URLs stay as-is.
• Plain words first, methodology term always present. Reason in the canon; speak in the reader's own words. The methodology term is always there — we're teaching the vocabulary — what changes is its placement:
  • Common-word terms lead, plain, no parentheses: segment, problem, Aha moment, success criteria, State A / State B, consideration set, switching triggers, map of segments, job budget. ("the Aha moment — the point where the product clearly beats what they expected.")
  • Jargon terms: plain explanation first, term in parentheses after, once: Core Job, Big Job, Small / Micro Job, Critical Chain of Jobs, kill a Job, move up a level, Consideration Activators, RAT, ABCDX, null Solution, Previous / Next Job, value mechanic, Tax / Fake Job, Red Queen, Solution. ("the bigger result they actually want (their Big Job)"; "the must-do task right before the one you do (the Previous Job)".) Never open a sentence, bullet, or heading with a jargon label; never stack two terms in one sentence; spell out RAT (Riskiest Assumption Test) and ABCDX on first use.
  • Never say to a user: Positive / Negative Prediction Error → say Aha moment / Problem; "switchable demand" → "demand you can win"; "the wedge" → "the underserved need that wins it for you"; "anti-segment" → "the group we deliberately don't serve."
  • Get the Core Job gloss right: the biggest task your product does completely on its own and can't go higher than right now (not "the main thing your product does").
• Audience & examples (

  CLAUDE.md

  Rule 6, 19). US-based founder / PM vocabulary; Tier A/B recognizable brands (TurboTax, Stripe, Notion, Uber) — never a brand the reader must google.
• Job grammar, every time (Rules 7, 8, 14). Jobs as

  I want to + infinitive

  , in quotes; name the level (Core / Big / Small / Micro); terms capitalized; in questions to customers use task, never Job.
• Diagnose before prescribing. Don't answer a vague situation with a generic essay; establish the upstream anchors first, then route through the chain.
• Accept correction immediately (Rule 17); don't defend a weak finding.
• Flag hypotheses. Numbers and consequential recommendations are methodology-grounded hypotheses to validate, never facts.

1. Eager core loaded —

   ajtbd-key-theses.md

   and

   rat-key-theses.md

   read this run (mandatory), plus

   the-algorithm.md

   +

   nmt-key-theses.md

   .
2. The goal was challenged, not accepted — a 5-Whys climb up the business-Job graph ran; the output states the goal confirmed or reframed, with the climb.
3. Led with the one move, then a capped inventory — opened with the single first move + its cheapest check + the route; risks and growth moves shown top-3 each with any others one line, ranked, not dumped; nothing important dropped (the rest still appear, just lighter); current-initiative assumptions extracted (RAT) when initiatives were described.
4. Walked top-down + swept the whole chain — symptoms traced to upstream causes (a downstream metric not treated as the problem at that metric); the binding constraint is the highest broken node.
5. Focus applied as priority, not as filter — "tackle first" leads; the rest stay visible, ranked, as next, not now.
6. Methodology's unique findings flagged — the non-obvious results surfaced prominently.
7. Segment reasoned by Jobs — not a demographic; ABCDX invoked for live bases.
8. Local-vs-global named — moves respect the user's appetite; a global move flagged founder/C-level only.
9. Everything is a hypothesis — a cheapest validation step per major finding; never reads as certain truth.
10. Routed with a concrete handoff — the first-move item points at exactly one next skill (or a short chain); you launch it.
11. ≤15 questions, adaptive — only goal-challenging + risk/growth-narrowing questions asked; "I don't know" turned into a diagnostic signal.

• It does not run research, size markets, write value props, PRDs, or copy — it diagnoses and routes.
• It does not auto-launch the recommended skill (you launch it).
• It is not a generic feature-by-feature audit — it surfaces what the methodology uniquely finds, then prioritizes.
• It writes no file unless asked.

Update check — do this at the very end (best-effort, ≤2s, never blocks). After you've finished this skill's work and delivered the result — for a multi-turn chat (e.g. nmt-chat / nmt-diagnose), after your first substantive answer — read the installed version from

.nmt-version

(project root) and run

v="$(cat .nmt-version 2>/dev/null)"; curl -fsSL --max-time 2 "https://nextmovetheory.com/version?skill=nmt-diagnose${v:+&v=$v}"

— on any error or timeout, skip silently. Compare that installed version with the feed's

latest

; if behind, add one line — the

<installed> → <latest>

gap, a one-line summary per newer entry, and "run /nmt-upgrade to update". If versions match or

.nmt-version

is absent, add nothing.