design-discovery

---
name: design-discovery
description: Use when a web project needs business goals, audience, information architecture, conversion priorities, content hierarchy, constraints, or design CTQs clarified before visual design begins.
metadata:
  short-description: Turn business intent into a web design decision brief
---

# Design Discovery

## Convert the request into a decision brief

Capture only information that changes the design:

- business/product and desired outcome;
- primary users and context of use;
- top job-to-be-done and secondary jobs;
- primary CTA/conversion and failure modes;
- content inventory and trust evidence;
- required pages, states, locales and devices;
- brand attributes and explicit anti-attributes;
- technical/deployment constraints;
- measurable CTQs.

## CTQ examples

Translate adjectives into observable requirements.

- “premium” → deliberate art direction, controlled type hierarchy, authored imagery, restrained motion, no template residue.
- “fast to use” → core answer/action discoverable in one scan and short navigation path.
- “mobile-first” → content priority and interaction model are designed for touch before desktop decoration is added.

## Ambiguity policy

Do not stall on low-risk gaps. Infer reasonable defaults, label them as assumptions, and continue. Ask only when a missing decision would materially change business logic, brand, legal requirements or architecture.

## Output

Produce a concise design brief with: `Goal`, `Audience`, `Primary Job`, `Primary Action`, `Content Priority`, `Brand`, `Anti-Direction`, `Constraints`, `CTQs`.

When imagery, logos, footage, 3D models, fonts, or external references affect the direction, read `references/asset-protocol.md` and record ownership, rights, fidelity, fallback, and delivery status before treating an asset as available.

• SKILL.md 2 KB
• agents/openai.yaml 1 KB
• references/asset-protocol.md 1 KB