gds-ux

Plan game UX, UI, and HUD design specifications as DESIGN.md and EXPERIENCE.md documents.

Updated Sep 3, 2026
One-click install
npx skills add https://github.com/watchthelight/shatterfish --skill gds-ux-watchthelight
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gds-ux
Source: https://github.com/watchthelight/shatterfish/tree/main/.claude/skills/gds-ux
Command: npx skills add https://github.com/watchthelight/shatterfish --skill gds-ux-watchthelight

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve? Game teams often jump into building UI and HUDs without a written contract for how the game should look and behave, causing inconsistent screens, rework, and misalignment between design and engineering. This Skill runs a structured elicitation workflow with the developer and produces two canonical specification documents that downstream architecture and story work can consume. ## Core Features & Use Cases - Dual specification output: Produces DESIGN.md (visual identity tokens per the Google Labs design.md spec) and EXPERIENCE.md (information architecture, HUD behavior, input schemes, game feel, accessibility, player journeys) that cross-reference each other by token name. - Guided discovery with creative tools: Elicits decisions from the developer without imposing taste, rendering HTML color themes, design directions, Excalidraw wireframes, and key-screen mockups on demand. - Validation and review: Offers an opt-in reviewer gate with parallel subagent lenses and a rubric walker that checks flow, token, component, and state coverage, emitting HTML and Markdown validation reports. - Use Case: A developer says "help me plan the game UX" for a roguelike; the Skill interviews them about platforms, input schemes, and HUD diegesis, then distills a final DESIGN.md and EXPERIENCE.md ready for the game architecture skill. ## Quick Start Ask the assistant to help you plan the game UX and HUD design for your game, and answer its discovery questions about platforms, inputs, and player journeys.

Frequently Asked Questions about gds-ux

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a game UX design specification with this Skill?

Invoke the Skill with a request like "lets create game UX" or "design the HUD". It runs a discovery interview about platforms, input schemes, and player journeys, then distills the decisions into DESIGN.md and EXPERIENCE.md files in a run folder.

What is the difference between DESIGN.md and EXPERIENCE.md?

DESIGN.md owns how the game looks: color, typography, spacing, and component tokens following the Google Labs design.md spec. EXPERIENCE.md owns how it works: information architecture, HUD behavior, states, input schemes, accessibility, and key flows, referencing DESIGN.md tokens by name.

Can the Skill validate an existing UX specification?

Yes. The Validate intent runs a rubric walker that checks flow coverage, token completeness, component and state coverage, and shape fit, plus optional reviewer lenses. It writes both validation-report.html and validation-report.md with findings grouped by severity.

Does the Skill support headless or automated runs?

Yes. When invoked headless it follows references/headless.md: it skips greetings and questions, infers intent from provided inputs, and ends with a JSON payload matching the schemas in assets/headless-schemas.md, including status, assumptions, and open questions.

What visual artifacts can the Skill generate during UX planning?

It can render HTML color theme comparisons, visual design directions, Excalidraw IA and flow wireframes, and 1:1 key-screen HTML mockups. Working artifacts land in .working/ and selected ones are promoted to mockups/ or wireframes/ at finalization.

When should I use a design handoff instead of the built-in workflow?

Use a design handoff when an external tool like Google Stitch should own the visual design work. The Skill assembles a producer-shaped prompt from captured discovery, you run the external tool, and save its outputs back into the workspace for EXPERIENCE.md to follow via Update mode.