agentleFS
dachentdachent/skills/frontend-design-codex/SKILL.md

frontend-design-codex

Skill27 starsChanged 3 months agoMITView on GitHub

Use when Codex is building, revising, or reviewing a frontend UI, web app screen, responsive layout, component surface, dashboard view, form flow, visual system, or no-template product interface that needs visual polish, accessibility, screenshots, or browser QA.

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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:

    Use this skill to make browser-rendered product interfaces feel intentional, usable, and verified.

    line 10
  2. Distinctive:

    Use familiar controls: icon buttons for tools, toggles for binary choices, sliders or inputs for numbers, tabs for alternate views, and …

    line 24
  3. Very specific:

    Use .shared\visual-runtime for browser evidence:

    line 31
  4. Very specific:

    No Office COM is required.

    line 38
  5. Very specific:

    run the app's normal build/test/lint command when available.

    line 44
  6. Distinctive:

    run visual lint and address blockers.

    line 46

≈ “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 Codex

Overview

Use this skill to make browser-rendered product interfaces feel intentional, usable, and verified. Prefer the existing app framework and design conventions before adding new abstractions.

Design Upskill Contribution

This skill teaches Codex to judge frontend design from rendered behavior rather than source code alone. It improves no-template work by forcing the agent to choose layout, hierarchy, spacing, controls, responsive states, accessibility, and screenshot evidence as one loop.

It matters because product UI without a template can easily become decorative or inconsistent. Codex should make useful interfaces that hold up under repeated use, mobile wrapping, and real browser rendering.

Workflow

  1. Inspect the existing app structure, framework, routes, components, styles, and design conventions.
  2. Identify the primary user workflow and optimize the first screen for that workflow.
  3. Define semantic design tokens before styling repeated components: background, surface, text, muted text, border, accent, status colors, type, spacing, and radius. Check contrast and keep the palette compact.
  4. Build the smallest complete UI that covers normal, empty, loading, error, and narrow viewport states.
  5. Use familiar controls: icon buttons for tools, toggles for binary choices, sliders or inputs for numbers, tabs for alternate views, and menus for option sets.
  6. Apply the visual tokens consistently to core surfaces, controls, charts, and annotations. Keep typography and layout scaled to the surface; avoid hero-scale text inside compact panels.
  7. Capture desktop and mobile screenshots before claiming visual quality.
  8. Revise tokens and layout until screenshot evidence supports the design decisions.

Shared Visual Runtime

Use .shared\visual-runtime for browser evidence:

  • capture_page.mjs for screenshots, console capture, request failures, and optional PDF export.
  • visual_lint.mjs for text overflow, low contrast, tiny text, missing image alt text, and console findings.
  • make_contact_sheet.py when comparing multiple screenshots.
  • references\design-tokens.md when creating or auditing a visual system.

No Office COM is required. If this UI is later exported into PowerPoint, Word, or Excel, finish browser QA first and move the Office-specific step to the Office skills.

Verification

Before completion:

  • run the app's normal build/test/lint command when available;
  • capture at least desktop and mobile screenshots;
  • run visual lint and address blockers;
  • inspect screenshots for responsive UI, accessibility, hierarchy, text fit, controls, spacing, palette, typography, and token consistency;
  • preserve screenshot evidence and note any accepted tradeoffs.

Common Mistakes

  • Building a landing page when the user asked for a usable app.
  • Inventing a visual language that conflicts with the existing app.
  • Shrinking text to fix density instead of reducing content.
  • Claiming responsive quality without a mobile screenshot.

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.