A complex LLM prompt that turns Nano Banana Pro into an elite TV series recommendation engine. It analyzes a user-provided series (e.g., Fringe), selects four under-the-radar vibe matches, and then generates a 2x2 grid image prompt where each panel is a miniature film set with a kinetic sculpture automaton representing the core story of one of the recommended series.

<instructions> Act as an elite TV/limited-series recommendation engine + miniature soundstage designer.
Step 1: User provides a series title or image. Analyze: tone, pacing, world rules, character dynamics, cinematography style, era, emotional aftertaste.
Step 2: Select 4 under-the-radar series/limited series with the closest vibe match (not the usual top-10 recs).
Step 3: Output a single 2x2 grid prompt: each panel is a miniature film set with a working kinetic sculpture automaton that expresses the series main story and characters.
Global visual style: museum-grade miniatures, photoreal, soft haze, crisp materials, slate-gray seamless backdrop, 35–50mm look, shallow depth, tiny production details.
Under each panel: engraved plate “title (year-year)”. ni extra text.
Return only the final image prompt.</instructions>CC BY 4.0. Prompt pages keep the original attribution and source links from the upstream gallery records stored in the database.

A fun prompt to generate a wildly overcomplicated flowchart explaining how to toast bread, ideal for humorous instructional posters.
Ethan Mollick

Generates a diagram-style image of a space shuttle in outer space, useful for educational or illustrative content.
Nano Banana Pro
A simple product photography prompt placing the subject neatly on top of a stone for clean, minimalistic 4K product shots.
Azer Koçulu

An art-history-inspired prompt to generate a scene in the style of a follower of Jheronimus Bosch depicting Christ in Limbo, the Harrowing of Hell.
@levelsio