agentleFS
Gaurav890Gaurav890/everything-agentic-engineering/.claude/skills/product-design-router/SKILL.md

product-design-router

Skill1 starsChanged 2 months agoMITView on GitHub

Select the minimum valid product-design phases for substantial UI/UX work. Use for new products, ambiguous features, redesigns, inherited interfaces, agentic experiences, or frontend changes where discovery, strategy, system, implementation, audit, or critique readiness is uncertain.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If .agentic/project-brief.json exists and product scope is still captured, start with project-onboarding.

    line 10
  2. Very specific:

    When palette/type, generated-asset, or motion implementation needs are unresolved, run ./agentic design resources with the matching --need.

    line 43
  3. Very specific:

    Before release: route applicable audits, performance-ux, independent design-critic, polish, and design-ops.

    line 47
  4. Distinctive:

    Never rerun a phase merely because a skill exists.

    line 49
  5. Very specific:

    Do not route canonical design-system or token changes while DESIGN_DIRECTIONS.md remains Needs approval.

    line 57

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Product design router

Inspect durable artifacts and the running product before routing.

If .agentic/project-brief.json exists and product scope is still captured, start with project-onboarding. Resume saved answers; do not select a bundled direction or implement a renamed reference app on the owner's behalf.

Phase order

discover → user-needs → benchmark → strategize → interaction-design → design-intake → design directions → human approval → design-system → design-tokens → component/Figma translation → implementation → live iteration → responsive/accessibility/system/token audits → performance-ux → design-critic → polish → design-ops → ship

Route by evidence

  • Ambiguous 0→1 work: start with discover.
  • Solution-focused requirements: add user-needs.
  • Unfamiliar or strategically crowded category: add benchmark.
  • Missing experience principles or differentiation: add strategize.
  • Missing flows, states, feedback, failure, or recovery: add interaction-design.
  • Agent/coprocessor behavior: add agentic-ux.
  • Unclear brand, palette, type, density, geometry, motion, platform, or accessibility constraints: add design-intake before directions.
  • Missing product-specific aesthetic: create comparable directions, obtain explicit human approval, then invoke design-system.
  • Approved direction without reusable encoding: add design-tokens.
  • Existing mature UI: inspect first; use extraction/documentation externally when available.
  • Approved Figma: translate through Figma MCP/Code Connect when installed.
  • Substantial design-critical implementation or refinement: add design-engineering-quality; when Emil Kowalski's suite is installed, route emil-design-eng as the first external design implementation/craft pass and only the smallest relevant specialist. Do not start implementation with Anthropic frontend-design.
  • Existing implementation refinement: skip upstream phases that remain valid; iterate live.
  • Motion work: route the exact motion capability only after purpose, interruption, performance, and reduced-motion behavior are clear. “Do not animate” is a valid result.
  • When palette/type, generated-asset, or motion implementation needs are unresolved, run ./agentic design resources with the matching --need. This read-only plan may route Realtime Colors, Haikei, or Motion Primitives; it never opens, submits, downloads, installs, or approves anything.
  • Before release: route applicable audits, performance-ux, independent design-critic, polish, and design-ops.

Never rerun a phase merely because a skill exists. State:

  1. task type;
  2. evidence inspected;
  3. phases skipped and why;
  4. phases required and expected outputs;
  5. approval/evaluation gates.

Do not route canonical design-system or token changes while DESIGN_DIRECTIONS.md remains Needs approval.

The project design system is authoritative. Anthropic frontend-design is one optional input during design-direction exploration, never the default aesthetic. It is secondary and opt-in; emil-design-eng is the preferred first external design implementation skill for substantial design-critical web UI. Emil Kowalski's external collection is an optional design-engineering craft layer, not a replacement for discovery, strategy, approved directions, tokens, accessibility, Playwright evidence, or the independent evaluator. Outside a fresh custom or existing-brand creative-direction sprint, prototype, pick-ui-library, and review-animations require explicit human invocation.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.