Aura-skill is an AI orchestration methodology that controls design intent at the prompt layer. It enforces four weighted axes across every output: Attitude, Utility, Refinement, and Aesthetics.
Not a style guide. Not a mood board. A standard with a reason behind every rule.
Load a skill into your agent, coding assistant, or system prompt and watch the difference between AI that produces and AI that decides. Taste is passive. Aura is active. Every skill here is defensible. You can explain it, test it, and teach it.
Every Aura skill must define four weighted control dials.
These are not suggestions. They are execution constraints.
Attitude: 0 — 10
Utility: 0 — 10
Refinement: 0 — 10
Aesthetics: 0 — 10
Example:
Attitude: 8
Utility: 9
Refinement: 7
Aesthetics: 6
AI must bias decisions toward higher-weighted axes.
If Attitude is high → bold decisions If Utility is high → function-first design If Refinement is high → polish + precision If Aesthetics is high → visual dominance
The dial values change behavior, not just description.
Every Aura skill must follow this structure:
# Aura Skill Name
## Dial Weights
Attitude: X
Utility: X
Refinement: X
Aesthetics: X
## Design Intent
What this skill is trying to enforce.
## Behavioral Rules
Hard rules AI must follow.
## Decision Bias
How dial weights affect decisions.
## Anti-Patterns
What this skill rejects.
## Output Requirements
What the final output must include.
AI should not proceed if dial weights are missing.
Attitude: 9
Utility: 6
Refinement: 7
Aesthetics: 8
This produces:
- Bold layouts
- Distinct visual hierarchy
- Strong personality
- Clean but expressive UI
Different weights produce different results.
Aura is controlled variation, not randomness.
| Skill | Enforces |
|---|---|
| aura-core | Design attitude engine — taste as a trainable discipline |
| aura-output | Anti-repetition system — precision as a habit |
| aura-redesign | Redesign audit layer — evolution over revolution |
| aura-brand | Brand hierarchy rules — distinctiveness over decoration |
| aura-product | Product UX discipline — inevitable over intuitive |
| aura-motion | Motion intentionality — meaning over movement |
| aura-editorial | Editorial authority — structure over style |
| aura-commerce | Friction elimination — confidence over convenience |
| aura-localbiz | Local trust architecture — authenticity over polish |
Verification tools. Run after applying a skill to confirm the work is complete and original.
| Playbook | Purpose |
|---|---|
| anti-cliche-library | Lazy-pattern rejection index |
| state-coverage-checklist | Completeness verification system |
| originality-enforcement | The "first prove it's new" protocol |
| hierarchy-rules | Visual priority framework |
Reference any skill directly in your AI tool, agent pipeline, or system prompt:
https://raw.githubusercontent.com/moswek/aura-skill/main/skills/aura-core
Recommended flow:
1. Load aura-core
2. Set dial weights
3. Apply domain skill
4. Run playbook
5. Generate output
6. Apply aura-output
Example:
Use aura-core
Attitude: 8
Utility: 9
Refinement: 7
Aesthetics: 6
Now build a premium SaaS dashboard.
If you have a discipline that belongs here, bring it. Fork the repo, create a branch, open a PR. One rule: state the standard, explain the reason, do not hedge. Read CONTRIBUTING.md before submitting.
All contributions must include dial weights.
MIT — Moses Wekesa, 2026