agentleFS
ytvee-devytvee-dev/webdev-agent-kit/skills/frontend-design-intelligence/SKILL.md

frontend-design-intelligence

Skill7 starsChanged 21 days agoApache-2.0View on GitHub

Review inspected designs or briefs for product-flow gaps and ask evidence-based user questions. Ground UI direction in audience, patterns, and UX risks; excludes code and tool setup.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 8 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create a compact design intelligence brief before visual direction when a UI task needs product-category grounding, page-pattern choice, …

    line 35
  2. Very specific:

    Use after prompt intent routing when the task is Standard Workflow or Deep Workflow and one or more are true:

    line 41
  3. Distinctive:

    Do not use for Fast Lookup, tiny edits, purely technical bugfixes, code implementation, live Figma inspection, package installation, or new …

    line 51
  4. Very specific:

    Read common/prompt-intent-routing-rules.md when workflow level is unclear.

    line 56
  5. Distinctive:

    Read relevant project overlays when present.

    line 57
  6. Distinctive:

    Read only relevant references from this skill:

    line 58
  7. Distinctive:

    Read supplied screenshots, briefs, inspect notes, or existing route files only when they materially affect the brief.

    line 62
  8. Distinctive:

    Must not run design generators or external CLIs.

    line 70
  9. Distinctive:

    Must not create or edit app source code.

    line 71
  10. Very specific:

    Must not create a global design system unless explicitly requested, and then only inside .agents/project/.

    line 72
  11. Distinctive:

    Every question identifies a real component/flow, the evidence gap, and user impact.

    line 118

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Frontend Design Intelligence

Purpose

Create a compact design intelligence brief before visual direction when a UI task needs product-category grounding, page-pattern choice, density decisions, or domain-specific anti-pattern checks.

This skill adapts design-system reasoning into the WebDev Agent Kit model without becoming a UI framework, CLI generator, or package installer.

When To Use

Use after prompt intent routing when the task is Standard Workflow or Deep Workflow and one or more are true:

  • product type, page type, audience, or user job is vague;
  • the user asks for visual style, landing structure, dashboard structure, or distinctive UI direction;
  • the page needs visual variance, motion, or density decisions;
  • inspected designs leave transitions, forms, states, or product behavior unclear;
  • frontend-design-director needs better grounding before writing a Design Direction Contract.

When Not To Use

Do not use for Fast Lookup, tiny edits, purely technical bugfixes, code implementation, live Figma inspection, package installation, or new styling systems.

Required Context

  1. Read AGENTS.md.
  2. Read common/prompt-intent-routing-rules.md when workflow level is unclear.
  3. Read relevant project overlays when present.
  4. Read only relevant references from this skill:
    • references/product-pattern-matrix.md;
    • references/design-dials.md;
    • references/product-anti-patterns.md.
  5. Read supplied screenshots, briefs, inspect notes, or existing route files only when they materially affect the brief.
  6. Read skills/design-screenshot-spec/references/product-behavior-review.md when reviewing flows or unresolved product/design decisions. If supplied visual sources have not been inspected, use design-screenshot-spec first.

Tool Contract

  • May read project overlays and relevant source files.
  • Must not run design generators or external CLIs.
  • Must not create or edit app source code.
  • Must not create a global design system unless explicitly requested, and then only inside .agents/project/**.
  • Delegate live source acquisition to design-screenshot-spec; consume its component evidence and report missing coverage before product review.

Workflow

  1. Identify product category, audience, page type, and single job.
  2. Separate confirmed facts from unknowns. Inspect the available evidence before asking the user; never fill missing product intent with an assumed default.
  3. Map the observed user journey and component states. Apply the product review reference to ask about unresolved transitions, forms, feedback, persistence, responsive behavior, and motion, with evidence and tradeoffs.
  4. Propose a page pattern and design dials only where choices are needed. Keep proposals distinct from confirmed constraints and obtain the user's decisions.
  5. Identify domain UX risks.
  6. Identify product-specific anti-patterns and generic template risks.
  7. Produce a compact Design Intelligence Brief.
  8. Hand off to frontend-design-director when visual direction is needed.

Output Contract

Final response: return only facts that affect the user's understanding, confidence, or next action. Omit empty fields and workflow narration.

Product Category
Audience
Page Type
Single Job
Recommended Page Pattern
Design Dials
Domain UX Risks
Product-Specific Anti-Patterns
Useful Subject Materials
Design Direction Handoff
Product Questions And Decisions
Blocked Dependent Scope

Validation Gates

  • The brief must not prescribe a UI library, styling system, package, or animation dependency.
  • The recommendation must be a starting hypothesis, not a rigid template.
  • The page pattern must connect to the user job.
  • Design dials must explain restraint or intensity.
  • Unspecified product/design choices remain open until answered; recommendations and existing-code behavior are not substitutes for a user decision.
  • Every question identifies a real component/flow, the evidence gap, and user impact. Do not re-ask settled facts or stop at a fixed total question count.
  • Durable records belong in local project overlays, not reusable bundle docs.

Trigger Evals

Should trigger:

  • "Ground this analytics dashboard before choosing its visual direction."
  • "Choose product patterns and design dials for this AI workflow."
  • "Review these inspected screens and ask what is missing from the form flow."

Should not trigger:

  • "Change this button color."
  • "Implement this completed Design Direction Contract."

Reference Map

  • references/product-pattern-matrix.md
  • skills/design-screenshot-spec/references/product-behavior-review.md
  • references/design-dials.md
  • references/product-anti-patterns.md
  • skills/frontend-design-director/SKILL.md
  • common/design-quality-rubric.md
  • common/anti-template-defaults.md
  • common/ui-ux-priority-checklist.md

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.