name-product

Imported from wfukatsu/nexus-architect (skills/product/name-product/SKILL.md). Install upstream with npx skills add wfukatsu/nexus-architect --skill name-product. Copyright stays with the author.

Product Naming (Acronym / Backronym)

Desired Outcome

Produce a product name that is spelled in the Latin alphabet and in which every letter is the first letter of an English word, so the name doubles as a phrase describing the product. Deliver:

• Product Name — reports/00_core/product-name.md
  • A word bank: English candidate words grouped by initial letter, drawn from the product's vision, values, differentiators, and domain (each word tagged with the theme it expresses)
  • Candidate names (default 5): each is a short pronounceable string plus its full expansion — one English word per letter — and a one-line read of what the expansion asserts
  • A screening table scoring each candidate on the naming criteria (distinct, short, pronounceable, appropriate, expandable, protectable)
  • A shortlist of 3 and one recommended name with rationale traced to VIS- IDs
  • Availability checks listed as Open Questions — never asserted as cleared
  • Each candidate and the recommendation carry a NAM- ID

The name is not decoration: it flows into the PR-FAQ heading, the positioning canvas, and every UI mock, so it must survive the same value story those artifacts tell.

Invocation

/product:name-product [target] [--input=<file|dir>]... [--count=N] [--style=acronym|initialism|hybrid] [--seed=<letters|word>] [--auto] [--lang=ja|en]

Decision Criteria

• The constraint is absolute: every letter of the final name maps to exactly one English word, and the per-letter words together read as a coherent phrase about the product. A candidate with a broken, forced, or nonsensical expansion is rejected, however nice it sounds.
• Pronounceable and short (acronym style): aim for 4–6 letters, ≤3 syllables, one obvious reading. initialism style may be shorter (2–4 letters) but must still expand.
• Appropriate, not generic: the expansion should describe this product's differentiation, not a category truism ("Fast Reliable Efficient System" describes everything → reject).
• Grounded, never fabricated: expansion words come from the product's own vision/values/ positioning vocabulary. Do not claim a name, domain, or trademark is available — availability is verified externally and logged to Open Questions.
• Stop condition: --count candidates exist, each with a valid full expansion and a screening score; a shortlist of 3 and one recommendation with rationale are written.

Prerequisites

Process

1. Read context — load --input, the vision/values, positioning, personas, scope, and work/traceability.json. Extract the load-bearing value keywords and differentiators, each with its VIS-/POS- source.
2. Elicit (gap-driven, skip on --auto) — confirm tone (technical vs. friendly), any letters/ words the user wants included or avoided, and language of the name (names stay Latin-alphabet even when --lang=ja).
3. Build the word bank — for each theme keyword, list strong English words and record each word's initial letter and the theme it expresses. This is the raw material for expansions.
4. Generate candidates — apply @rules/product/naming-frameworks.md. Work both directions: forward (assemble initials from value words into a pronounceable string) and backward (pick a pronounceable string, then fit one English word per letter from the bank). Honor --seed. Produce --count candidates, each with a complete letter-by-letter expansion.
5. Screen — score each candidate on the six naming criteria; drop any with a forced or generic expansion. Note pronunciation and syllable count.
6. Availability — list the exact checks to run (trademark class, .com/domain, app-store, collision with the word bank's own competitors) as Open Questions with status external — they need outside verification, so they are recorded with an owner rather than put to the user (@rules/open-questions.md §2); never mark them cleared.
7. Shortlist & recommend — pick 3, then 1, with rationale that cites the VIS- IDs the expansion satisfies. Assign NAM- IDs.
8. Append traceability — for each NAM- node add {id, type:"name", title, skill:"name-product", source_file, upstream:[VIS-…, POS-…]} to work/traceability.json.
9. Record — write reports/00_core/product-name.md; append the recommendation to work/context.md; ask remaining unknowns and log only what stays open (@rules/open-questions.md) — availability checks always stay open as external.

Output

reports/00_core/product-name.md — word bank, candidate table with full expansions, screening scores, shortlist, one recommended name, a NAM- ID table with an Upstream column, and an Open Questions block for availability. The name string itself is always Latin-alphabet regardless of output_language; surrounding prose follows the configured language.

Reference Materials

Related Skills