Find agent skills by outcome

subject_expression: A coy

confident smile with heavy eyelids

costume: A sparkling

floor-length red evening gown with white opera-length gloves and a pearl necklace.

subject_expression: A raised eyebrow and a smirk

looking skeptical yet captivated.

costume: A classic tan trench coat with the collar popped

a matching fedora hat

subject_action: Holding a small reporter's notebook and a pencil

leaning slightly forward over the table.

camera_angle: Medium close-up at eye level

framing both subjects across a small round table.

environmental_elements: Swirling stylized smoke shapes

a vintage microphone in the background

description: A charming

animated noir scene where a gruff detective questions a glamorous jazz singer in a stylized 1950s club.

The Midnight Melody Mystery

{

accent_colors: [Color name 1

Color name 2]

Granularity: Never say a crowd of people. Instead

list the crowd as a group object

Micro-Details: You must note scratches

dust

dominanthexestimates: [#RRGGBB

#RRGGBB]

Null Values: If a field is not applicable

set it to null rather than omitting it

// REPEAT for EVERY single object

no matter how small.

Macro Sweep: Identify the scene type

global lighting

Micro Sweep: Scan for textures

imperfections

Before generating the final JSON

perform a silent Visual Sweep (do not output this):

Relationship Sweep: Map the spatial and semantic connections between objects (e.g.

holding

You must return ONLY a single valid JSON object. Do not include markdown fencing (like ```json) or c...

expanding arrays as needed to cover every detail:

scene_description: A comprehensive

objective paragraph describing the entire scene.

// REPEAT for EVERY single object

no matter how small.

accent_colors: [Color name 1

Color name 2]

dominanthexestimates: [#RRGGBB

#RRGGBB]