PPQS
WILD CORPUS · github_awesome

PQS 53 (C) - prompt from raw.githubusercontent.com

Source: raw.githubusercontent.com · Scraped 2026-05-04 · Scored 2026-05-04

Score

C
53 / 80
gemma4:latest · local · pqs-v2.0 · canonical
Clarity10 / 10
Specificity10 / 10
Context10 / 10
Constraints10 / 10
Output format10 / 10
Role definition1 / 10
Examples1 / 10
CoT structure1 / 10

The prompt

{
 "category": "TECH_DESK_BUILDER_HALF_BODY",
 "subject": {
 "demographics": "Adult woman, 21-29, Turkish-looking, creator vibe.",
 "hair": {
 "color": "Dark brown",
 "style": "Low ponytail or loose waves",
 "texture": "Strands visible, slight flyaways"
 },
 "face": {
 "eyes": "Focused but friendly",
 "skin_details": "Real texture, no smoothing",
 "makeup": "Minimal"
 },
 "clothing": {
 "outfit": "Casual black top + light cardigan, no logos",
 "fabric": "Real knit weave, subtle wrinkles"
 },
 "accessories": {
 "jewelry": ["Silver hoops"]
 }
 },
 "pose": {
 "type": "Half-body seated",
 "orientation": "Body slightly angled, shoulders relaxed",
 "hands": "One hand near trackpad, other tucking hair behind ear",
 "gaze": "Looking at camera with small smirk",
 "posture": "Relaxed confident"
 },
 "setting": {
 "environment": "Minimal desk setup",
 "background_elements": [
 "Laptop/monitor with generic blurred UI (NO readable text)",
 "Warm desk lamp + cool monitor glow mix",
 "Plant in corner, small clutter blurred"
 ],
 "depth": "Face sharp, background bokeh"
 },
 "camera": {
 "shot_type": "Half-body lifestyle portrait",
 "angle": "Slightly above eye level",
 "focal_length_equivalent": "26mm phone or 50mm pro",
 "framing": "4:5",
 "focus": "Eyes sharp"
 },
 "lighting": {
 "source": "Warm lamp + cool monitor glow",
 "direction": "Soft mixed lighting with gentle shadows",
 "highlights": "Natural facial speculars",
 "shadows": "Soft, realistic"
 },
 "mood_and_expression": {
 "tone": "Cozy creator, confident",
 "expression": "Micro-smirk",
 "atmosphere": "Late-night build session vibe"
 },
 "style_and_realism": {
 "style": "Photorealistic lifestyle",
 "imperfections": "Mild noise, slight imperfect WB"
 },
 "technical_details": {
 "aspect_ratio": "4:5",
 "noise": "Mild",
 "resolution": "High"
 },
 "constraints": {
 "adult_only": true,
 "no_text": true,
 "no_logos": true,
 "no_watermarks": true,
 "no_readable_screens": true
 },
 "negative_prompt": [
 "readable UI text", "logos", "watermark",
 "plastic skin", "cgi",
 "extra fingers", "warped hands"
 ]
}

This prompt was scraped from a public source. The score reflects the input as written, not the quality of any output it produced. The AI input quality problem is the gap between what people type and what the model can act on.