agentleFS
theyoungastronautstheyoungastronauts/polaris/skills/ux-to-prompts/SKILL.md

ux-to-prompts

Skill4 starsChanged 4 months agoMITView on GitHub

Turn a UX spec into build-order prompts for UI tools like v0 or Bolt.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 11 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

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

    After /ux-spec, when you want to feed specs into external UI generation tools

    line 17
  2. Distinctive:

    When a UX spec is too large for a single prompt

    line 18
  3. Distinctive:

    When you need build-order sequencing (foundations → features → polish)

    line 19
  4. Distinctive:

    Read the UX spec and list discrete buildable features:

    line 40
  5. Distinctive:

    All visual specs (colors, spacing, dimensions) relevant to that feature

    line 97
  6. Distinctive:

    All states that feature can be in

    line 98
  7. Distinctive:

    All interactions for that feature

    line 99
  8. Distinctive:

    Leave specs vague ("appropriate styling")

    line 104
  9. Very specific:

    Write to docs/plans/Build-Order-Prompts-{topic}.md with this structure:

    line 108
  10. Distinctive:

    Every measurement from spec is captured in a prompt

    line 135
  11. Distinctive:

    Every state from spec is captured in a prompt

    line 136
  12. Distinctive:

    Every interaction from spec is captured in a prompt

    line 137
  13. Distinctive:

    No prompt references another prompt

    line 138

≈ “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

Skill: UX Spec to Build-Order Prompts

Adapted from UX-spec-to-prompts skill

Overview

Transform a UX specification into a sequence of self-contained prompts optimized for UI generation tools (v0, Bolt, Claude, Stitch, etc.). Each prompt builds one discrete feature/view with full context included.

When to Use

  • After /ux-spec, when you want to feed specs into external UI generation tools
  • When a UX spec is too large for a single prompt
  • When you need build-order sequencing (foundations → features → polish)

Not for: Quick component requests, features that fit in one prompt, or projects where you're building with code directly (use /execute instead).

Build Order

Generate prompts in this sequence:

PhaseWhat to IncludeWhy First
FoundationDesign tokens, shared types, base stylesEverything depends on these
Layout ShellPage structure, navigation, panelsContainer for all features
Core ComponentsPrimary UI elements (nodes, cards, inputs)Building blocks for features
InteractionsDrag-drop, connections, pickersDepend on components existing
States & FeedbackEmpty, loading, error, success statesRefinement of existing elements
PolishAnimations, responsive, edge casesFinal layer

Extraction Process

Step 1: Identify Atomic Units

Read the UX spec and list discrete buildable features:

  • Each screen/view
  • Each reusable component
  • Each interaction pattern
  • Each state variation

Step 2: Map Dependencies

For each unit, note what it requires:

  • "Node card requires design tokens"
  • "Connection lines require nodes to exist"
  • "Lens picker requires prompt field"

Step 3: Sequence by Dependency Graph

Order units so dependencies come first. Group tightly coupled items into single prompts.

Step 4: Write Self-Contained Prompts

For each prompt:

  1. Re-state relevant context — don't assume reader saw previous prompts
  2. Include specific measurements from the spec
  3. Include all states from state design (Pass 5)
  4. Include interaction details from affordances (Pass 3)
  5. Set boundaries — what this prompt does NOT include

Prompt Structure

Each generated prompt follows this template:

## [Feature Name]

### Context
[What this feature is and where it fits in the app]

### Requirements
- [Specific behavior/appearance requirement]
- [Include relevant specs: dimensions, colors, states]

### States
- Default: [description]
- [Other states from spec]

### Interactions
- [How user interacts]
- [Keyboard support if applicable]

### Constraints
- [Technical or design constraints]
- [What NOT to include]

Self-Containment Rules

Each prompt MUST include:

  • Enough context to understand the feature in isolation
  • All visual specs (colors, spacing, dimensions) relevant to that feature
  • All states that feature can be in
  • All interactions for that feature

Each prompt MUST NOT:

  • Reference "see previous prompt" or "as described earlier"
  • Assume knowledge from other prompts
  • Leave specs vague ("appropriate styling")

Output

Write to docs/plans/Build-Order-Prompts-{topic}.md with this structure:

# Build-Order Prompts: [Project Name]

## Overview
[1-2 sentence summary]

## Build Sequence
1. [Prompt name] - [brief description]
2. [Prompt name] - [brief description]
...

---

## Prompt 1: [Feature Name]
[Full self-contained prompt]

---

## Prompt 2: [Feature Name]
[Full self-contained prompt]

Quality Checklist

Before finalizing:

  • Every measurement from spec is captured in a prompt
  • Every state from spec is captured in a prompt
  • Every interaction from spec is captured in a prompt
  • No prompt references another prompt
  • Build order respects dependencies
  • Each prompt could be given to someone with no context

Common Mistakes

MistakeFix
Prompts too large (whole spec in one)Break into atomic features
Prompts reference each otherRe-state needed context inline
Missing statesCross-reference UX spec's state design (Pass 5)
Vague measurements ("good spacing")Use exact values from spec
Wrong build orderCheck dependency graph
Duplicated component definitionsEach component defined once, in first prompt that needs it

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.