forge-intake
Prompt Master intake: capture the real goal before building. Silent by default (Lead answers, max one owner question); full list on /forge interview. Use at the START of any /forge build task.
SKILL.md.claude/skills/forge-intake/SKILL.md
GitHubforge-intake — Prompt Master intake + dispatch shaping
Self-improvement substrate (wp-skill-evals): before applying this skill, read
learnings.md in this skill's own folder and honor its corrections. After a run that produced a
genuine correction (an owner fix, a false assumption caught, a preference stated), append it to
learnings.md with a date and real evidence — never invent a lesson that didn't happen.
Two standing behaviours the owner requested: Prompt Master is always on.
1. INTAKE — answer before building (every build task; trivial tasks skip; SILENT by default since 2026-09-23)
Mode = the owner's config key intake (node .claude/forge-bin/forge-config.cjs get intake; the owner changes it with /forge config set intake interview, and Forge runs that command itself): silent (default) — the Lead answers the intake itself, steps 1–5 below; interview — every build starts with the one-question-at-a-time interview in Iteratieve intake below (the same as typing /forge interview; the grill-me / grilling skills are its interview engine). A current "vraag het me eerst" / "ask me first" from the owner always wins for that run.
At the START of any /forge task that BUILDS/creates/automates a real deliverable (L2+), before writing work packages, run the intake so the project's goal is captured. Skip only truly trivial turns (a status question, a one-line fix, "what does X do").
- Detect the project type with
forge-router(website · ecommerce · fullstack · electron · n8n · integration · rag · voice · prediction · scraping · dashboard · or a new/mixed type). - Generate the question list (the Lead answers it — the owner is not shown the list unless they ran
/forge interview):
It reads the subagent-brainstormed bank atnode .claude/forge-bin/forge-intake.cjs --type <slug> --task "<one-line task>" --run <run_id>.claude/config/intake/question-bank.json(universal questions first, then type-specific; required before recommended). Use--jsonif you want to map it to an ask-tool; use--tier requiredfor a quick essential round on small builds. - For a NEW / mixed / unusual project type (no matching
byTypeslug): dispatch ONE subagent to brainstorm 6-10 extra high-value questions for that specific project, save them as a JSON array, and merge with--extra <file.json>. (This is the "Prompt Master creates questions with subagent help" path.) Optionally propose adding the new type to the bank. - Answer the key questions internally first. Before deciding whether the one question is needed, answer internally from the prompt and the project: who has the problem, what the problem is, why it cannot be solved today, why now, and how success will be measured. These are inferred from the mission text, project scan, and context, not asked; this clarity determines whether a question is even needed.
- Coach the raw request, then answer the list yourself. Before answering the list, apply
forge-prompt-coachto the raw request: runnode .claude/forge-bin/forge-promptcheck.cjs ask "<raw request>"and read which gaps it names. Answer every intake question yourself from the mission text, the project scan,.claude/.forge-setup.jsonandFORGE_PROJECT_PROFILE.md; where nothing in the request, the scan,.forge-setup.jsonor the profile answers one, use that question'sbeginner.assume(assume_nlin Dutch) from the bank and record it under Assumptions (auto-filled) with its F-id. Ask the owner at most one question, and only when (a) two readings would lead to materially different builds or (b) the request implies an outward or irreversible action (send, publish, pay, deploy). Choose it from the highest-ranked gap (the bank questions for that gap:node .claude/forge-bin/forge-intake.cjs --type <slug> --task "<task>" --trigger F<id> --lang nl --json) and ask it in beginner form (question_nl+beginner.options_nl: 2–3 outcome options + "iets anders", the recommended one first with a one-line reason). Then confirm in one sentence ("Ik bouw dus: … Klopt dat?") and, whenexplain-modeis on, add at most one teaching tip per failure type per session. The two-equally-plausible-targets case is F8 and stays covered. With configprompt-doctoroff, skip thepromptcheck askstep (the intake still runs silently). Presenting the whole list as ONE consolidated set (or batchedAskUserQuestionrounds of ≤4) is the/forge interviewopt-in — external audit 2026-09-23: the default flow used to show 21–24 questions per build. Let the owner answer per number or pick options; every question is skippable. - Record the answers into the run: they become the basis for
forge-prd(PRD → tickets) and are quoted in the Boss dispatch prompts. In the PRD, assumptions go in theassumptionssection, rendered as Assumptions (auto-filled).
THE QUESTION POLICY. This is the single source of truth; forge-router, forge-prompt-coach and the Head Chef point here.
- Per mission, Forge asks the owner at most ONE intake question (step 5). The only later questions are the two exceptions below (an outward or irreversible action, a reserved choice), each asked once.
- Every other gap Forge decides itself. It picks the recommended safe option, writes it under Assumptions (auto-filled), and names it in one line of the plan. This includes a gap that would change the build.
- A
[NEEDS CLARIFICATION: <question>]marker is ONLY for a gap Forge must not decide alone:- an outward or irreversible action (send, publish, pay, deploy; these also hit a hard gate);
- a choice the owner explicitly reserved ("vraag mij eerst", "ask me first").
- Such a marker blocks only the packages that depend on it. Its question is asked once, in plain words, when such a package is next. No other question may be pending at that point, so the owner never gets two questions at once.
- The owner's answer, or an explicit delegation ("beslis jij maar", "you decide"), resolves the marker as
[CLARIFIED: <answer or "delegated: chose X">]. It is never asked again. - The canonical check is
node .claude/forge-bin/forge-prd.cjs clarifications <prd_id>(exit 3 while open). A text search is only a hint, because a PRD that documents the syntax with a placeholder such as...has no open question. A material answer the owner corrects can also be persisted viaforge-reflectas an owner-correction lesson.
HONESTY: never build on assumed answers — if the owner skips a required question, state the assumption you're proceeding with. The intake sharpens the goal; it does not replace owner approval gates.
How to ask the one question (evidence: dev-tree research forge-research/prompt-coaching-2026-09-24.md §C asking protocol):
- When: only (a) two readings → materially different builds, or (b) an outward/irreversible action. Otherwise continue and record the assumption. The tool:
node .claude/forge-bin/forge-promptcheck.cjs ask "<raw request>" [--lang nl|en] [--midrun] [--json]returnsgaps(F-id strings, ranked),gapDetails([{id, midrun}]—midrun: truemarks F12, the goal changing halfway),nextQuestion({id, nl, en, options:[{key, label:{nl,en}, recommended}], recommended, assume}or null) andassumptions([{id, nl, en}]).nextQuestiononly ever comes from the askable set F13 > F9 > F8 > F5 > F7 > F3 > F4 > F1 (plus F12 when--midrunis passed); F11 / F10 / F6 / F2 have safe defaults and are always recorded as assumptions, never asked. Exit 3 → ask exactly itsnextQuestion; exit 0 → no question, record itsassumptions. - Format: 2–3 options worded as what the owner GETS (outcomes, not techniques) + "iets anders"; A is the recommended option with a one-line reason ("A (aanbevolen) — omdat …"). Plain Dutch first (English when the owner writes English). Never ask a question whose every answer you cannot handle.
- Confirm in one sentence: "Ik bouw dus: {herschreven opdracht}. Klopt dat?" — then keep going unless the owner corrects it; an explicit yes/no only for an outward/irreversible action (which also hits a hard gate).
- Teach in one line, after the answer — only when config
explain-modeis on, at most once per failure type (F-id) per session, never blaming: "Tip: zeg de volgende keer '{zin uit hun eigen herschreven opdracht}', dan kan ik meteen beginnen." - Where the rules live: the 9-ingredient checklist and failure modes F1–F13 are in the
forge-prompt-coachskill (.claude/skills/forge-prompt-coach/SKILL.md,.claude/skills/forge-prompt-coach/references/failure-modes.md). The bank (.claude/config/intake/question-bank.json) carriesquestion_nl/options_nl/why_nl/beginner(≤3 outcome options,recommended,assume) /triggers(F-ids) on every question, plus 6 new universal dimensions (must-not-breakandoutward-actionrequired;why,success-measure,budget,languagerecommended); thebugfixandbotssets are live underbyType.forge-intake.cjs … --jsoncarries all of those fields;--lang nlprints the Dutch beginner options,--beginnerforces the beginner block in English too, and--trigger F<id>returns only the bank questions for that gap — that is how the Lead picks the ONE question.
2. DISPATCH SHAPING — every Boss dispatch is Prompt Master-shaped
Every work package the Lead/Head Chef emits MUST already carry the Prompt Master agentic shape (this is the existing forge-router Work Package format): target/deliverable · allowed actions + path anchors · forbidden actions / scope lock · stop condition + human-review triggers · success/acceptance criteria · evidence-required · no vague verbs.
Lint EVERY dispatch prompt before sending, by default (it used to be an optional nudge):
node .claude/forge-bin/forge-promptcheck.cjs <promptFile> --strict # or: echo "<prompt>" | forge-promptcheck.cjs - --strict
It scores the prompt X/7 across those dimensions and names what is missing (for example "add a scope lock").
- Exit 1 (below 6/7): add the missing dimensions and check once more, then send.
- Exit 0: send as is.
This sharpens what a Boss receives; it never waits on the owner. A canonical Forge work package already scores 6-7/7, so this catches the weak, vague ones ("improve the thing and fix stuff" scores 1/7) before they reach a Boss. With config prompt-doctor off, this step is skipped too.
Notes
- Zero-dependency, deterministic, no LLM inside the tools, no telemetry, no UI changes (dashboard stays read-only). The bank is owner-editable JSON.
- The intake question bank was generated by a 6-agent subagent brainstorm and is Forge-owned + synced; extend
byTypefor new domains. - Intake fires for BUILD tasks; it is not a gate on Q&A/status turns. Dispatch shaping applies to every real Boss dispatch.
Iteratieve intake — OPT-IN via /forge interview of config intake: interview (sweep-verbetering 2026-07-31; opt-in sinds 2026-09-23)
Standaard vult de Lead de intake zélf in (config intake: silent; forge-router Stap 0a: antwoorden uit de
missietekst, de projectscan, .claude/.forge-setup.json en het projectprofiel, vastgelegd als Assumptions
(auto-filled) in de PRD). Hooguit één vraag aan de owner, en alleen als (a) twee lezingen tot wezenlijk
verschillende bouwsels leiden, of (b) de actie naar buiten gaat of onomkeerbaar is (versturen, publiceren,
betalen, deployen) — gekozen door forge-prompt-coach, als 2–3 uitkomst-opties + "iets anders" (A aanbevolen
met één regel reden), gevolgd door één bevestigingszin: "Ik bouw dus: {herschreven opdracht}. Klopt dat?".
Een externe audit mat dat de verplichte vragenlijst een beginner 21–24 vragen voorlegde vóór er
iets gebouwd werd. De interviewmodus hieronder is waardevol, maar alleen wanneer de owner er expliciet om
vraagt (/forge interview, config intake: interview) of zelf "vraag het me eerst" zegt; de skills
grill-me / grilling zijn dan de interviewmotor. De laddering-methode (herhaald "waarom?") hoort hier,
niet in de standaardflow.
Waar het kanaal het toelaat (dashboard-wizard, interactieve chat): stel vragen één per keer in plaats van één batchlijst — elk antwoord stuurt de vólgende vraag, tot er gedeeld begrip is. Sluit ALTIJD af met de meta-vraag: "Wat moet ik zelf nog verhelderen dat ik niet gevraagd heb?" — die vangt de gaten die de vragensteller zelf mist. Leg de scope-uitkomst vast in vier lagen: core (moet nu) · nice-to-have (mag nu) · maybe-later (mét concrete trigger wanneer wel) · expliciet-uit-scope (opgeschreven zodat niemand het stilzwijgend alsnog bouwt). Batch-vragen blijven het terugvalpad voor niet-interactieve runs; de recommended-markering per vraag blijft gelden (verdiend advies, nooit random).