High-level product definition and UX discovery skill. Runs seven gates to determine whether a product is worth building, who it's for, what experience should exist, what the MVP boundary is, and what UI/UX direction should guide implementation. Produces .pm/stable/ files and initializes .pm/runtime/state.yaml. Use when: product idea, "worth building", "product definition", "define this product", user journey, UX principles, UI direction, MVP boundary.
Instalación
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
High-level product definition and UX discovery skill. Runs seven gates to determine whether a product is worth building, who it's for, what experience should exist, what the MVP boundary is, and what UI/UX direction should guide implementation. Produces .pm/stable/ files and initializes .pm/runtime/state.yaml. Use when: product idea, "worth building", "product definition", "define this product", user journey, UX principles, UI direction, MVP boundary.
harness-grill-product
Seven-gate product discovery. Each gate must pass before the next begins. The goal is to freeze a product contract that Supervisor can use to drive delivery.
Before Starting
Check for existing .pm/stable/ files. If they exist, ask whether to update or start fresh.
If .pm/ directory does not exist, create it with subdirectories stable/, runtime/, design/, design/generated-concepts/.
Read any existing README, product notes, or design references the user provides.
Assess product maturity: ask the user which stage applies:
Pre-product (idea only): Run all 7 gates. This is the default.
Has users (launched, has real users): Skip Gate 1 demand validation. Focus on Gate 4-6 for next iteration.
Has paying customers: Skip Gates 1-2. Focus on Gate 4 (MVP wedge for next version) and Gate 5-6.
Hackathon/learning project: Switch to Builder mode (see below).
Mode Selection
At the start, ask the user:
"What kind of project is this?"
A) Startup — I want to build something people will use or pay for. Scrutinize the idea hard.
B) Builder — I'm building to learn, experiment, or for fun. Help me think bigger, but keep it fun.
Startup mode (default): Runs all applicable gates with full diagnostic pressure. Push back on weak evidence. Be the critical investor.
Builder mode: Replace Gates 1-2 with generative questions:
"What's the coolest version of this you can imagine?"
"What would make someone say 'whoa' when they see this?"
"If you had no constraints, what would this become?"
Keep Gates 3-7 for structural discipline (you still need to know who it's for and what it looks like).
Compressed mode: For quick dogfood or when the user says "fast" / "quick pass":
Ask only 1-2 highest-value questions per gate
Generate v0 stable files immediately after all gates touched
Mark ux_depth: light and evidence_status: weak in state.yaml
User can re-run individual gates later to strengthen specific areas
Gates
Gate
Goal
Output
1. Demand Reality
Determine if the idea corresponds to real demand
evidence.md, product.md sections
2. Product Reframe
Find the 10x version hiding behind the obvious request
product.md, value-proposition.md, decisions.md
3. User & Scenario Specificity
Define first users and core scenario with desperate specificity