Dashboard/Workflow Canvas

Agentic Creative Platform · The production engine

One brief in. On-brand variations out — at scale.

Build a workflow once — wired to your brand strategy, colors, briefs, and research — then run it repeatedly at scale. Move fast in a guided App, or go deep on the full node Canvas. Same engine, same project underneath. Agencies go from brief to campaign-ready creative in minutes, not days.

Brief
AI variations
Staged assets
Published campaign
Live performance

One canvas, two ways in

Move fast in an App, or go deep on the Canvas

Every StyleForge App is a friendly, guided surface over the same full node engine. Race through a guided flow, or flip to Canvas Mode for node-level control — it's the same project underneath, and you can switch anytime without duplicating a thing.

Guided Apps

A numbered step panel: write a brief, pick a Style, add references, press Generate. Variations stream into a filmstrip — keep the winners, publish, done. The node complexity is handled for you.

The full Canvas

Snap nodes together into any pipeline — chain models, wire in brand context, composite and color-grade, branch outputs. Total control for power users, with everything pre-wired and ready to generate.

Open a project (yours + your team's)Load a saved templateOpen a shared template by IDStart a guided workflowPick a platform pipelineBlank canvas

The engine

Brand context + a brief → N × M on-brand variations

This is the difference. A powerful orchestrator node reads your brand strategy, colors, creative brief, and reference images, then writes N distinct creative directions — and generates M takes of each. Every one of the N × M results is a different idea rendered inside your brand's visual language, not a random spread. Author the N briefs yourself with a Variations Node, or let the orchestrator propose them.

Brand strategy
Brand colors
Creative brief
Variation briefs (N)
Reference images
Orchestratorreads brand + brief

N variations × M takes — all on-brand

A Style (archetype) sets the creative framework and its dials — e.g. Vary Copy — Lock Scene keeps your proven visual and tests only the message. Pick a Style, dial it in, set N and M, and run. Then iterate: feed a favorite back in as the reference and evolve it.

The powerful AI nodes

A roster of orchestrators, each purpose-built

These are the agentic nodes that do the heavy creative direction. Each one ingests your brand context and a brief, then drives a batch of on-brand outputs toward a specific goal.

Ad Variation Generator

Iterate a winning ad, reverse-engineer a competitor's layout into your brand, or factory net-new concepts — three modes, one run.

Winning Ad

Feed a proven ad and its live performance; it spins brand-consistent variations to amplify the winner.

Ad Hook Lines

Copy-led ad concepts where the hook, subline, and CTA are written and rendered as typography — many distinct angles at once.

Character Maker

Lock a character from one reference image, then generate a full sheet of poses that all keep the exact same look.

Angle Maker

Controlled angle progressions of any subject — ideal for A/B testing perspectives on people, products, and creative.

Cinematic Scene Director

Turns a brief into a multi-panel storyboard, then animates it into one continuous cinematic film.

UGC Actor Video

Describe an actor and pick a product; the video engine casts the actor and produces UGC-style videos in batches.

Klaviyo Universal Content

One brief → Hero, Product Grid, and Product Statement email blocks, each its own rail, published to Klaviyo.

POD Content Generator

Designs, blank mockups, UGC actor shots, and full brand colorways from one brief — composited and publish-ready.

Shopify Product & Theme Analyzer

Pull a product or theme, analyze it against your brand, and generate on-brand imagery straight back to the store.

YouTube Thumbnails

Click-worthy, brand-consistent thumbnail variations across Thumbnail, Shorts, A/B, Series, and Brand-channel modes.

Workflow Architect

Describe what you want in plain English and it builds the entire wired node graph on the canvas for you.

Or just describe it — the Workflow Architect builds it

Tell the Workflow Architect what you want in plain English — “build a UGC avatar video pipeline with greenscreen compositing” — and it assembles the entire wired node graph on the canvas for you, ready to run. Not a template picker: real nodes, really connected.

Guided Quick-Start Apps

Nine one-click apps over the same engine

Each app bundles an orchestrator, its inputs, and its publish targets into a guided flow — pick one, name the project, and you're generating. Most are agent-drivable too.

Ad Variator

Agent-ready

Ad variations

Winning Ad

Agent-ready

Amplify a winner

UGC Actor Video

Agent-ready

UGC actor videos

Cinematic Video

Agent-ready

Storyboard → film

Product Video

Agent-ready

Cinematic product video

Ad Hook Lines

Agent-ready

Copy-led ad concepts

Shopify Product Generator

Agent-ready

On-brand product imagery

Klaviyo Universal Content

Agent-ready

Publish-ready email blocks

POD Content Generator

Print-on-demand, composited

Open the canvas to launch an app

The node arsenal

150+ node types, organized the way the canvas is

Everything below is a node you drag onto the canvas and wire up — grouped exactly as the node sidebar groups them. A representative few per category; there's more inside each.

Agentic Orchestration

The powerful AI nodes that read your brand and direct the generation.

Ad Variation GeneratorWinning AdCharacter MakerAngle MakerCinematic Scene DirectorWorkflow ArchitectPrompt EnhancerImage & Video Analyzer+12 more

Image Models

15+ state-of-the-art image models, switchable from a single node.

Nano Banana ProFlux 2 ProImagen 4 UltraGPT ImageRecraft v4Ideogram v3SeeDream 5Qwen Image+6 more

Video Models

The frontier of text- and image-to-video, all from one node.

Veo 3Sora 2 ProKling 3Seedance 2Hailuo 02Wan 2.2Grok ImagineKling Omni+4 more

UGC Image & Video

Pre-wired UGC workflows — actors, greenscreen, and motion.

UGC Actor VideoUGC AvatarMotion Graphics+3 more

Compositors

Combine images and video — static and motion — into one output.

Canvas CompositorMulti-Layer Video CompositorVideo ConcatenatorImage Animation (Ken Burns)+2 more

Color & Tone

A professional, non-destructive color-grading suite.

LevelsCurvesLUT ColorExposurePhoto FilterImage Adjust

Editing

Transform and composite assets without leaving the canvas.

Inline Image EditorInline Video EditorImage UpscaleRemove BackgroundFocus EffectBrush EffectsSORN Scene Analysis+4 more

Audio & Music

Generate voiceover narration and original music on-canvas.

Audio NarrationMusic Generator

Input & Brand Context

Feed prompts, references, and your brand into any AI node.

Brand StrategyBrand ColorsBrand Kit AssetBrand TopicShopify ProductAd Campaign AssetVariationsList+6 more

Utility & Publish

Route finished work to your brand kit, campaigns, and store.

Save to Brand KitSave to CampaignSave to ShopifySave to KlaviyoEmail PreviewDownload+4 more

Every top model, one node, batch-ready

Switch models from a single node — no re-wiring. Connect a Variations Node and fire N independent generation jobs in one click, each combining your base prompt with a unique creative brief.

Image — 15+ models

Nano Banana ProFlux 2 ProImagen 4 UltraGPT ImageRecraft v4Ideogram v3SeeDream 5Lucid Origin (Leonardo)Qwen Image+ more

Video — 8+ models

Veo 3Sora 2 ProKling 3Seedance 2Hailuo 02Wan 2.2Grok ImagineKling Omni+ more

Exact poses & camera control

Direct the shot — not just the prompt

Text prompts can't reliably nail a pose or a camera angle. These nodes can. Set the 3D pose, the lens, the light, and the film stock, and the top models match it precisely.

Rodin 3D → pose

Generate a 3D model from an image, orbit to the exact angle, and Save View — that snapshot becomes a locked pose reference the top models match precisely.

Camera angle presets

Eye level, bird's eye, Dutch angle, over-the-shoulder — wired preset nodes that feed the prompt automatically.

Lens & depth of field

Wide, telephoto, macro, anamorphic + aperture and DoF presets for precise photographic control.

Lighting & film stock

Golden hour, studio strobe, neon, cinematic grain — swap one node to change the whole shoot aesthetic.

Finish it on the canvas

Pro color, editing & compositing — inline

Grade, retouch, composite, and animate without exporting to another tool. Non-destructive, chainable, and it writes straight back into the pipeline.

Color & Tone

Six non-destructive grading nodes — including LUT Color with film-stock emulation (Kodak Vision3, Fuji Eterna).

LevelsCurvesLUT ColorExposurePhoto FilterImage Adjust

Editing

Transform assets in place — a full image editor, a video editor, and one-click cleanups.

Inline Image EditorInline Video EditorImage UpscaleRemove BackgroundFocus EffectBrush Effects

Compositing

Stack layers with automatic greenscreen removal; embed a full design compositor with Smart Merge that survives regeneration.

Canvas CompositorMulti-Layer Video CompositorVideo ConcatenatorImage Animation

The full video pipeline lives on the canvas. Generate scenes → concatenate with professional transitions → composite actors, graphics, and background (auto greenscreen) → animate stills with Ken Burns motion → add narration + music → and use SORN scene analysis to break any video into editable per-scene prompts, swap one scene, and regenerate — all without leaving the workflow.

Pre-wired creative direction

Preset libraries do the heavy lifting

Drag a preset onto the canvas and it feeds controlled creative direction straight into your prompt rig. Curated libraries you can use as-is or fork into your own — this is where the volume lives.

Preset Lists

Source nodes of controlled creative direction — swap one to change the whole look.

Camera AnglesLens TypesApertureLighting MoodsFilm StockCompositionLight StreamsCast ShadowsFilm Grain+ more, incl. full fashion-attribute libraries

Preset Variations

Pre-populated variation sets that drive a whole batch in one click — 50+ across these themes.

Ad creative
HooksAnglesDemographicsCTAsEmotions+3 more
Product photography
AnglesContextLifestyleDetailSeasonal+3 more
Fashion editorial
DirectionSettingTreatmentNarrativeStyling+3 more
Compositional disruption
Crop energyFrame-breakingNegative spaceVisual tension+4 more
Camera & motion
MotionFocus playLight paintingLens character+4 more
Creative strategy
Brand voiceNarrative arcCultural momentEnergy level+4 more

200+ pre-wired solutions

Click a solution and the entire node setup drops onto a new project — fill in your content and run. Ad publishing, video production, UGC, design and more.

Guided, annotated workflows

Step-by-step labels on every node walk you through Ad Variations, Winning Ad, Hook Lines, Character Maker, Angle Maker, YouTube Thumbnails and more.

Templates, shared across the team

Save any workflow as a template — keep it private to your team or share it by ID. Open a teammate’s project and pick up right where they left off.

Wire your brand in once — every model stays on-brand

Connect your Brand Strategy, Colors, Assets, and Topics to any AI node with a single edge. Your voice, visual guidelines, palette, and research are injected into every generation in the pipeline — the AI isn’t guessing, it’s working from your actual brand.

Brand Strategy nodeBrand Colors nodeBrand Kit AssetsBrand Topics & ResearchUGC actorsLogos

Finished creative goes straight out. Route any output to your Brand Kit, or use Save to Campaign to push into StyleForge’s ad campaign platform — organize into groups and stage for Meta, Google PMax, and TikTok (all published paused, so nothing spends until you say so). Shopify and Klaviyo save nodes publish back to your store and email too.

Works inClaude.ai (web)Claude CoworkClaude Code— no API key needed.

Run the whole engine from a conversation

The same orchestrator engine you drive by hand, your connected agent can drive by chat. It picks the right app, wires in your brand context, sets the N × M archetype spread, fires the run, and streams the numbered results to a live preview — then iterates on the ones you like and publishes the keepers.

pick the app
wire brand context
set N × M + Style
fire the run
watch it stream
iterate a pick
save / publish

Spin a batch of on-brand ads

Ask Claude to

Make 6 on-brand ad variations for our new trail runner — vary the hook and the scene.

Direct the archetype spread

Ask Claude to

Use the Hero and Product Grid styles, 4 variations each, 2 takes per variation.

Watch it, then evolve it

Ask Claude to

I like image 3 — iterate on it with a bolder headline.

Restock a store or fill a campaign

Ask Claude to

Pull our best-seller and generate lifestyle shots, then publish them back to Shopify.

See all agent capabilities →·Connect Claude →