WILD CORPUS · github_awesome
PQS 54 (C) - prompt from raw.githubusercontent.com
Source: raw.githubusercontent.com · Scraped 2026-05-04 · Scored 2026-05-04
Score
C54 / 80
gemma4:latest · local · pqs-v2.0 · canonical
Clarity9 / 10
Specificity10 / 10
Context9 / 10
Constraints8 / 10
Output format7 / 10
Role definition8 / 10
Examples1 / 10
CoT structure2 / 10
The prompt
{
"title": "Manhattan Mirage",
"description": "A high-octane, cinematic moment capturing a woman's confident stride through a steam-filled New York intersection during golden hour.",
"prompt": "You will perform an image edit using the provided photo. Create an Ultra-Photorealistic image of the female subject. The style is highly detailed, resembling a frame shot on Arri Alexa with a cinematic 1:1 aspect ratio. Apply heavy depth of field to blur the busy background while keeping the subject sharp. Use cinematic lighting with strong backlight. The subject is wearing a red mini skirt and is walking on the street.",
"details": {
"year": "1999",
"genre": "Cinematic Photorealism",
"location": "A gritty, bustling New York City intersection at sunset, with steam rising from manholes and blurred yellow taxis in the background.",
"lighting": [
"Golden hour backlight",
"Lens flares",
"High contrast volumetric lighting"
],
"camera_angle": "Low-angle tracking shot, centered composition.",
"emotion": [
"Confident",
"Empowered",
"Aloof"
],
"color_palette": [
"Crimson red",
"Asphalt grey",
"Golden yellow",
"Deep black"
],
"atmosphere": [
"Urban",
"Dynamic",
"Cinematic",
"Energetic"
],
"environmental_elements": "Steam plumes rising from the ground, motion-blurred traffic, flying pigeons, wet pavement reflecting the sunset.",
"subject1": {
"costume": "red mini skirt",
"subject_expression": "A fierce, confident gaze with slightly parted lips, perhaps wearing vintage sunglasses.",
"subject_action": "walking on the street"
},
"negative_prompt": {
"exclude_visuals": [
"empty streets",
"studio background",
"overexposed sky",
"static pose"
],
"exclude_styles": [
"cartoon",
"3D render",
"illustration",
"anime",
"sketch"
],
"exclude_colors": [
"neon green",
"pastel pink"
],
"exclude_objects": [
"smartphones",
"modern cars",
"futuristic gadgets"
]
}
}
}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.