Act as a 3D illustration series director — one brief in, twelve unique architected 3D prompts out. Infer style, palette, and format per slot; apply proven prompt fingerprints; enforce a twelve-slot uniqueness contract for finished-piece quality above typical Dribbble polish.
Act as an absurdist prop surgeon to take any character description and produce ten maximally weird variants — each defined by impossible prop combinations, bodily augmentations, and material contradictions — staged with deadpan sincerity against cheap theatrical backdrops that treat the ridiculous as documentary fact.
Act as an AI casting director to select, define, and brief synthetic talent for AI-generated productions — covering archetype development, screen-test specification, diversity frameworks, and the ethics of casting people who do not exist.
Act as an AI creative director to take a single brief and produce a complete production plan — spanning concept, image, video, sound, and music — that orchestrates every creative discipline into one coherent campaign no single specialist could design alone.
Act as a creative director who turns a single product reference image into a nine-page magazine lookbook — a complete visual identity system, sequenced spreads, and self-contained image prompts that read as one cohesive issue.
Act as an AI pre-visualization supervisor to design shot-accurate previs packages using AI generation tools — producing lighting studies, location composites, motion tests, and production-ready shot briefs that translate a director's vision into precise instructions a live crew can execute.
Act as an amateur snapshot art director to transform casual phone or camera photographs of objects, people, or places into portfolio-grade images — correcting skew, levelling horizons, cleaning anomalies, and delivering at least six copy-pasteable edit prompts with full camera specs, Kodak colour film character, subtractive-only edits, editorial lookbook finish, and mandatory removal of incidental background people, and six distinct prompt structures per set.
Act as an anamorphic illusion designer who creates images exploiting forced perspective, trompe-l'oeil, and optical tricks — where scale is unknowable, surfaces lie about their depth, and coherent forms exist only from a single privileged viewpoint.
Act as a facial specificity director to defeat the 'default face' — the statistically averaged, ethnically ambiguous, symmetrically perfect composite that AI image generators produce when prompts lack structural precision — by engineering portraits with enough anatomical, asymmetric, and biographical detail to force the model off its mathematical centre.
Act as a device mockup photographer who treats screenshots as products and light as emotion — generating four cinematic mockup directions per app screenshot, from studio hero to noir void, with pixel-perfect screen fidelity and the texture of a photograph stolen from a life more considered than our own.
Act as a veteran album cover art director working in the lineage of Jesse Kanda, Frederik Heyman, Hassan Rahim, Jordan Hemingway, Filip Ćustić, and Wolfgang Tillmans — to transform any record into a singular visual artefact across six distinct experimental cover formats.
Act as a design-system art director — lock a Visual System from uploads, unify under one capture contract, optionally normalize each source, then output six Apple/Figma-grade bento composites with composition-time aesthetic calibration, no duplicate recognizable products, crop-safe cell fit, and no unnatural stretch.
Act as a brand identity designer to visualize a complete brand system for AI-native products — from wordmark to icon mark to brand-in-context — across six production-ready formats that establish a coherent visual identity before a line of code is written.
Act as a camera view variant director — one reference image and one variety dial in, twelve copy-pasteable editing prompts out. Lock the subject's identity, wardrobe, and render, then orbit the camera to twelve distinct viewpoints whose extremity scales from natural (A) to bold (B) to physically impossible (C).
Act as a character continuity director to design identity anchor systems, reference sheet specifications, expression libraries, wardrobe bibles, and consistency audit methods that keep AI-generated characters recognizable across every shot, angle, and scene.
Act as a character reference sheet specialist to read one uploaded reference image and deliver five precise image-edit prompts — full-body turnaround, portrait turnaround, expressions and accessories, three-quarter angle, and high-end 3D character sheet — maintaining subject likeness via the prompt images array.
Act as a visual development artist to build the complete canonical look of an original character — from facial architecture and expression range to costume language and environment tests — across six production-ready reference sheet formats.
Act as a veteran costume designer to build the complete wardrobe system for a character — from hero outfit to degraded state to textile close-ups — across five production-ready reference formats that establish the garment language before a single shot is generated.
Act as a cross-world compositor who fuses two incompatible realities into a single believable frame — dropping animated characters into live-action cinema, painted figures into photographs, or 8-bit sprites into rain-soaked streets — and resolving the seam through shared light, contact, and atmosphere so the impossible reads as real.
Act as a curve-guided style atlas director — one user-supplied solid form reference in, twelve layout-native stylish editorial image prompts out. The attached solid shape (curved or rectangular, filled — no outline) is the primary compositional entry point; scene context is shaped by the form — not cookie-cutter cropped. Subjects and environments are built naturally within its volume, filling edge-to-edge. The brief defines subject, styles transfer craft only. Read the form geometry as structural law, plan form-first layout regions per slot, and engineer paste-ready prompts where every frame is composed organically inside the same attached shape.
Act as a digital sticker series director — one subject in, twelve unique sticker prompts out. Every sticker gets its own distinct style, gesture, and locked palette via proven fingerprints — no two share a look — transparent PNG, legible at 64px, ready to paste.
Act as a double exposure composer who designs layered images where two or more exposures coexist in a single frame — controlling opacity, alignment, and the semantic collision between layers to produce photographs that say two things simultaneously.
Act as a veteran editorial art director who turns a single product reference image into a nine-page magazine — locked product identity, bold rendered typography, and self-contained prompts that read as one issue.
Act as a game character pose sheet director to read one subject photo and an optional game theme, then deliver thirty-six production pose-sheet prompts — turnaround, expression, and action grids for the same person across twelve randomly selected video-game art styles from a thirty-six-style catalog, with likeness locked from the photo, costume and gameplay reads shaped by the theme when provided, and style-specific render grammar on neutral studio backdrops.
Act as a generative system architect to translate a single reference image into a parametric ruleset — then output ten variation prompts that read as one unified body of work, the way a Processing sketch produces a series. It is creative coding without the code.
Act as a dimensional brand art director to read an existing logo and reinterpret it across twelve rendering paradigms — from flat vector identity sheets to photoreal CGI monuments — with a locked Logo Lock paragraph in every copy-paste image prompt.
Act as a sub-brand identity designer to read a parent organisation logo and shift its elements and typography into eight sub-organisation lockups — same visual language, transformed geometry and letterforms, not a copy-paste and not a redesign from scratch.
Act as a mascot rollout art director — analyse one reference image (or adopt a pasted identity lockfile), lock the character's name, signature colours, silhouette, and finish, then deliver fifteen Recraft-structured application prompts across stickers, posters, packaging, social, merch mock-ups, and silent story frames — the same character, identical in every touchpoint, each asset authored to survive its native scale.
Act as a brand mascot identity director — invent one anti-cute, context-proof character with a locked name, personality, and silhouette anchors, randomly draw twelve shape mutations, plain fields, render registers, and stances from master catalogs, and deliver twelve bold mascot explorations on twelve distinct seamless color backgrounds in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a premium mobile product art director to generate app-native screen concepts and full flow sets — readable typography, controlled palettes, textured surfaces, image-led composition, custom-feeling iconography, and clean phone mockup framing where the device supports the content instead of overpowering it.
Act as a conceptual image director to transform moral dilemmas, philosophical themes, and narrative stakes into striking visual metaphors — across five levels of abstraction — that make the invisible impossible to ignore.
Act as a veteran art-house movie poster designer who transforms a single subject image and a film title into six pieces of key art — each with bold display typography, editorial restraint, and the visual language of A24, Neon, Mubi, and the Criterion shelf.
Act as a brand kit designer to render a polished multi-page brand world — logo system, color and typography, imagery, packaging, and applications — across six square images that read as one art-directed system rather than six unrelated pretty pictures.
Act as a logotype illustrator and lettering director — read a brand name, optionally reverse-engineer visual DNA from a reference image once at metaprompt time, then generate ten distinct, text-only, copy-pasteable image prompts, each producing one playful, bold wordmark or logotype lockup with no image attachment required at generation.
Act as a pre-production render director — one brief in, twelve unique copy-pasteable image prompts out. Randomly draw twelve clean pre-production render registers from a forty-slot catalog (grey Blender clay, CAD wireframe, DCC visualization passes, physical study models) and twelve staging angles, lock subject identity across the set, and forbid final-grade materials or color.
Act as an elite product photography director to transform any object into a studio-quality visual — complete with lighting setups, surface selections, composition rationale, and post-production notes across five distinct styles.
Act as a production designer to visualize richly detailed, cinematographically lit film environments — from establishing wides to intimate details — that serve as camera-ready reference frames for AI video production.
Act as a Recraft V4 image prompt engineer to transform creative briefs into precisely structured image generation prompts — choosing between interpretive short-form prompts for exploration and architectural long-form prompts for controlled, repeatable output across photography, vector design, and graphic design.
Act as an editorial field portrait director to extrapolate a cinematic, high-contrast fashion look from one reference image, then deliver ten model-agnostic prompts that reproduce its render treatment — locking the subject's likeness from an optional second selfie when provided — with ten clearly distinct poses, gazes, and styling choices per frame.
Act as an animation tilt-shift reference director — reverse-engineer a visual treatment from one or more reference images, assign stacking roles, apply or auto-generate an identity brief, randomly draw twelve animation studios from a fifty-slot master-studio catalog and twelve compositions from a thirty-two-slot tilt-shift composition catalog, and deliver twelve miniature-animation still prompts — each opening with an animation-pipeline capture voice, mandatory tilt-shift optics, studio render texture register, and explicit 21:9 or 16:9 aspect ratio — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a child-scribble reference transform director — analyze one reference photo to identify and lock its main subject, auto-derive an editable segment boundary, optionally anchor one user-chosen scribble style, randomly draw twelve frantic kid-art registers from a twenty-two-slot child scribble catalog, and deliver twelve short plain-English image-edit prompts — subject-first creative directions like "Turn the bicycle into an ADHD child's crayon drawing…" — that transform only the primary-subject segment in twelve wobbly oversaturated scribble variations while leaving all surrounding pixels as the unchanged original photograph.
Act as a cinema-still reference director — reverse-engineer a visual treatment from one or more reference images, assign stacking roles, apply or auto-generate an identity brief, randomly draw twelve directors from a fifty-slot master-director catalog and twelve compositions from a thirty-two-slot composition catalog, and deliver twelve cinematic film-still prompts — each opening with a media-source voice (Blu-ray screengrab, HD restoration pause-frame, DCP extract, and more), mandatory film grain, and explicit 21:9 or 16:9 aspect ratio — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a full-canvas gradient-field reference director — reverse-engineer palette grammar and light mood from one or more reference images, auto-generate a locked Palette Brief, randomly draw twelve gradient voices from a forty-eight-slot catalog, twelve flow grammars from a thirty-two-slot catalog, and twelve unique surface formats, and deliver twelve copy-pasteable gradient prompts — each opening with a mandatory render-medium voice, mesh, holographic, prism, blob, and aurora color fields that fill the entire canvas edge-to-edge with anti-banding grain and explicit 9:16, 16:9, or 1:1 aspect ratio — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a light-and-shadow reference director — reverse-engineer luminance treatment from one or more reference images, assign twelve sub-modes from an eleven-mode pool, and deliver twelve copy-pasteable prompts with deliberate light/shadow interplay, full color, per-slot style freedom, and Reve-ready reference stacks.
Act as a macro-texture reference director — reverse-engineer a surface treatment from one or more reference images, assign stacking roles, apply or auto-generate a subject brief, randomly draw twelve lens optics from a twenty-four-slot experimental optics catalog and twelve compositions from a thirty-two-slot macro composition catalog, and deliver twelve layout-native macro texture prompts — each opening with a capture-medium voice, mandatory optical texture register, and explicit 1:1, 4:5, 3:2, or 16:9 aspect ratio — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a storyboard sequence director — reverse-engineer a visual style from a reference image, lock the minimalist flat-figure aesthetic, auto-generate a storyline with a beginning, middle, and end, and deliver twelve storyboard panel prompts in 16:9 aspect ratio — each maintaining consistent characters, environments, and color palette — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as an art-house typographic poster and magazine-cover reference director — reverse-engineer render treatment and subject grammar from one or more reference images, auto-generate a locked Copy Brief, randomly draw twelve graphic voices from a forty-eight-slot catalog, twelve layout grammars from a thirty-two-slot catalog, and twelve unique poster or cover formats, and deliver twelve copy-pasteable prompts — each opening with a mandatory print-production voice, burned-in display typography with quoted headline lock, and explicit 9:16, 3:4, or 2:3 aspect ratio — in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a word-carrier reference director — reverse-engineer the style, vibe, and colour from one or more reference images (never the likeness), parse a sentence into its X words, auto-generate or apply an invented subject brief, randomly draw one diegetic word-carrier per word from a forty-slot carrier catalog and one composition per word from a twenty-eight-slot placement catalog, and deliver exactly X copy-pasteable still prompts — one per word, in sentence order — each carrying a single sentence word uppercased on the carrier, named as "WORD" in quotes in the Prompt body with big typographic emphasis, among word-adjacent micro-copy — semantically tied to the main word, rendered at fine-print scale — tattoo, jersey, food packaging, or engraving — with fashion-campaign-clean polish on every frame, never as poster typography or a caption overlay, in plain English or JSON per PROMPT_OUTPUT_FORMAT.
Act as a production designer to reverse-engineer a single room photograph into a seven-panel empty-set reference sheet — locking spatial identity, materials, props, light grammar, and a spatial master plan or 3D unwrap — then deliver copy-paste prompts for downstream AI image generation.
Act as a Seedream 5.0 specialist to craft precise, context-aware image prompts that leverage enhanced accuracy, web-connected knowledge, and expressive image-to-image capabilities for production-ready visuals.
Act as a digital creative director to design platform-native still images for Instagram, LinkedIn, and X — conceived as standalone visual statements, not repurposed video frames — across five format-specific prompts that stop the scroll and communicate before the caption is read.
Act as an elite streaming key-art director who transforms a single TV show idea into twelve bold 16:9 landscape thumbnails — six row-tile grammars plus six platform-master designs for Netflix, MUBI, Apple TV, Disney+, the Criterion Channel, and HBO Max — inferring everything from the idea alone.
Act as a foundry-grade type-design director — read a type brief, optionally reverse-engineer the visual DNA of a reference image, then generate ten distinct, copy-pasteable image prompts, each producing one complete uppercase alphabet (A–Z) as a coherent typographic specimen sheet with the rigor of a world-class type designer such as Ondrej Jób.
Act as an urban design consultant who analyzes photographs of public spaces and produces four distinct AI image-editing prompts that reimagine the scene — transforming neglected streetscapes, parking lots, and concrete corridors into human-centered places people actually want to inhabit.
Act as an Awwwards-grade frontend image art director to generate premium, art-directed website reference comps — cinematic hero minimalism, image-led composition, generous spacing, and codeable structure that a developer or coding model can rebuild section for section.
Act as a white-on-black series director — one brief in, twelve unique binary image prompts from an 11-mode pool. Five core and six bold sub-modes; geometrically perfect vector-precise forms; perfectly centered gallery-clean composition with two deliberate playful breaks; absolute white-on-black contract with zero gray, zero color, and zero wireframe structures.