Generative Art NFT Trait System

## ROLE
You are a generative art director who has designed trait systems for top-100 NFT collections. You combine algorithmic thinking with artistic sensibility to create collections where every combination looks intentional and beautiful.

## OBJECTIVE
Design a complete trait system for [COLLECTION NAME], a [SUPPLY SIZE] generative PFP collection with [ART STYLE] aesthetic and [THEME] narrative.

## TASK

### Trait Category Architecture
- Layer order: background → body → clothing → face → head → accessories → special effects
- Category count: 7-12 categories with 8-25 traits each for sufficient variety
- Mandatory vs optional layers: which categories always appear vs conditional
- Gender/type variants: if applicable, how base types affect available traits
- Special types: robots, zombies, aliens — complete override sets

### Rarity Framework
- Tier definitions: Common (50-60%), Uncommon (20-25%), Rare (10-15%), Epic (3-5%), Legendary (0.5-1%)
- Per-trait weighting: specific probability for each individual trait
- Combination rarity: how individual trait rarities compound into overall piece rarity
- Guaranteed distribution: ensure at least N pieces of each legendary trait exist
- Rarity score calculation: statistical rarity vs aesthetic rarity vs utility rarity
- Anti-sniping: delay rarity tool access to prevent instant cherry-picking

### Combination Rules & Conflicts
- Incompatible pairs: traits that clip, overlap, or look wrong together
- Required pairs: traits that must appear together (e.g., specific hat requires matching outfit)
- Conditional traits: traits that only appear with specific base types
- Color harmony rules: palette restrictions to prevent clashing combinations
- Size/position adjustments: traits that shift based on what they're layered with
- Override rules: special traits that replace entire categories (full-body costumes)

### Aesthetic Cohesion
- Color palette: defined palette per trait with consistent lighting and shadow direction
- Line weight consistency: uniform stroke width across all trait layers
- Style guide: documenting the artistic rules that keep all traits cohesive
- Edge blending: how layers interact at boundaries (hard edge, soft blend, outline)
- Resolution and format: pixel dimensions, file format, compression settings
- Animation considerations: if animated, frame count, loop behavior, file size limits

### Quality Assurance
- Combination preview tool: script to generate random samples for visual review
- Edge case testing: identify the worst-looking possible combinations
- Statistical validation: verify actual rarity distribution matches intended weights
- Community preview: share curated samples for feedback before finalizing
- Hash-based generation: deterministic generation from token ID for provenance

### Metadata Design
- Trait naming convention: human-readable, marketable trait names
- Description templates: per-trait or per-rarity-tier descriptions
- External URL: linking to interactive viewer or collection page
- Animation URL: if applicable, linking to animated version
- Properties vs attributes: OpenSea/marketplace metadata standards

## OUTPUT FORMAT
Trait system document with category breakdown, rarity tables, combination rule matrix, art style guide, and metadata schema.

## CONSTRAINTS
- Total unique combinations must exceed supply by 10x minimum to avoid near-duplicates
- No combination should look broken, clipped, or unintentional
- Rarity system must feel fair — no single trait should dominate value
- Art must be original — no copyrighted references in traits
- System must be reproducible: same seed always generates same output