# Lp Do Assessment 05 Name Selection > Produce a -naming-generation-spec.md for a business. Reads ASSESSMENT stage docs, extracts ICP, product, brand personality, competitive set, and any prior eliminated names, then writes a fully-structured agent-executable spec that generates 250 scored candidate names. Part 1 of a 4-part naming pipeline (spec → generate → RDAP batch check → rank). - Skill: `majiayu000/lp-do-assessment-05-name-selection` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds add majiayu000/lp-do-assessment-05-name-selection` - Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/lp-do-assessment-05-name-selection/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: DevOps & Infra - Author: majiayu000 (https://skillmd.com/u/majiayu000) - Updated: 2026-09-09 - Page: https://skillmd.com/skills/majiayu000/lp-do-assessment-05-name-selection --- # lp-do-assessment-05-name-selection — Name Selection (ASSESSMENT-05) Produces a `-naming-generation-spec.md` tailored to a specific business. The spec is the input to a name-generation agent that produces 250 scored candidates, which are then batch-checked for .com availability via RDAP and ranked. This skill replaced the former naming research prompt approach (obsolete since deep research tools cannot verify domain availability). **Upstream dependency:** Operates on ASSESSMENT-04 output. Requires at minimum: a problem statement, a product/option decision, and a brand personality sketch. Load: ../_shared/assessment/assessment-base-contract.md ## When to use Use when a brand name is open (not yet committed) and the business has sufficient ASSESSMENT stage context. Run on user instruction: `/lp-do-assessment-05-name-selection --business ` ## The four-part naming pipeline This skill produces Part 1 only. The full pipeline is: | Part | What | Who runs it | |------|------|-------------| | **1 — Spec** | This skill. Reads ASSESSMENT docs, writes `-naming-generation-spec.md` | This skill | | **2 — Generate** | Agent reads the spec and produces 250 scored candidate names | Spawn a general-purpose agent with the spec as input | | **3 — RDAP batch check** | Shell loop hits `https://rdap.verisign.com/com/v1/domain/.com` for all 250; 404 = available, 200 = taken | Bash tool | | **4 — Rank** | Filter to available names, sort by score, produce final shortlist | Agent or inline | ## Operating mode **SPEC AUTHORING ONLY** **Allowed:** read ASSESSMENT docs, synthesize context, write `-naming-generation-spec.md`. **Not allowed:** generate candidate names, check domains, make naming recommendations. ## Required inputs (pre-flight) Read and synthesise from: - `docs/business-os/strategy//-problem-statement.user.md` — falsifiable problem, user segments - `docs/business-os/strategy//s0c-option-select.user.md` — selected product option(s), price points - `docs/business-os/strategy//-brand-identity-dossier.user.md` — brand personality, positioning, visual direction - `docs/business-os/strategy//s0a-research-appendix.user.md` — segment truth, purchase triggers, competitive set - `docs/business-os/strategy//s1-readiness.user.md` — ICP confirmation, outcome contract Also check for prior naming rounds (any `*-naming-shortlist*.user.md` files under the strategy dir). If present, extract all eliminated names into the spec's elimination list. If any file is missing, note the gap in a preflight comment at the top of the spec but proceed with available context. Do not halt. ## Output Save to: ``` docs/business-os/strategy//assessment/naming-workbench/-naming-generation-spec.md ``` If the file already exists (a prior naming round was run), **update it in place**: add newly eliminated names to the elimination list, update the ICP section if it has changed, and increment the round note at the top. Do not overwrite a spec from scratch. ## Spec structure (mandatory sections) The output spec must contain all of the following sections. Sections §3, §4, and §5 are business-specific and must be populated from the ASSESSMENT docs. Sections §3.3 and §5 (generation patterns and output format) are universal infrastructure — copy them verbatim from the template below, do not rewrite. ### §1–§2: Brand Context and ICP Resonance Load: modules/part-1-2.md ### §3–§4: Scoring Rubric and Generation Patterns Load: modules/part-3-4.md ### §5–§6: Hard Blockers, Output Format, and Quality Gate Load: modules/part-5-6.md ## Completion message > "Naming generation spec ready: `docs/business-os/strategy//assessment/naming-workbench/-naming-generation-spec.md`. Next: spawn a general-purpose agent with this spec as input to generate 250 scored candidates. Then run the RDAP batch check. Then filter and rank." ## Integration **Upstream (ASSESSMENT-03):** Runs after `/lp-do-assessment-03-solution-selection` produces a product decision record. **Downstream — Part 2:** Spawn a general-purpose agent with the spec. Prompt: "Read `docs/business-os/strategy//assessment/naming-workbench/-naming-generation-spec.md` in full, then generate exactly 250 brand name candidates following the spec exactly. Save to `docs/business-os/strategy//assessment/naming-workbench/naming-candidates-.md`." **Downstream — Part 3 (RDAP batch):** Extract all names from the candidates file. Run: ```bash while IFS= read -r name; do status=$(curl -s -o /dev/null -w "%{http_code}" "https://rdap.verisign.com/com/v1/domain/${name}.com") echo "$status $name" done < names.txt ``` Filter to 404s. For Pattern D names, check the domain string, not the spoken name. **Downstream — Part 4 (rank):** Filter candidates table to domain-available names only. Sort by Score descending. Present top 20 as the working shortlist for operator review. Save to `docs/business-os/strategy//assessment/naming-workbench/naming-shortlist-.user.md`. **Downstream (ASSESSMENT-06):** `/lp-do-assessment-06-distribution-profiling` runs after the operator has confirmed a working name from the shortlist.