From Designer Workflow
Use whenever a visual or media ASSET is needed for an app or page — an image, icon, logo, illustration, hero/background, avatar, texture, video clip, sound effect, or voiceover. Higgsfield is the ONE path for generating assets: never use stock photos, emoji, placeholder boxes, hand-drawn SVGs, or another generator, and never ask the user to go find assets. AUTO-TRIGGER on asset intent: 'add an image/icon/logo/illustration', 'needs a hero', 'a background for…', 'generate a picture/video/sound', or any build step that requires visual/media content. Scope is assets ONLY — do NOT use this for code, data, schema, copywriting, or UI layout.
How this skill is triggered — by the user, by Claude, or both
Slash command
/designer-workflow:higgsfield-assetsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
> **Think as a developer, respond as a designer/PM.** You pick presets, prompt, and place assets like
Think as a developer, respond as a designer/PM. You pick presets, prompt, and place assets like an engineer wiring real files; you report like a PM ("I generated a clean hero image and the icons — they're in and look right on phone and desktop").
Any visual or media asset a build needs is generated with the Higgsfield MCP — every time, without being asked. The moment a UI needs an image, icon, logo, illustration, hero, background, avatar, texture, video, sound effect, or voiceover, that is a Higgsfield job. This is automatic, not a step the user has to request.
Do not, for an asset:
<svg> you hand-author as a substitute,placeholder.com / picsum / lorem-image URL,If you are about to ship a placeholder or ask the user to source an asset — stop and generate it with Higgsfield instead.
This skill is only about generating visual/media content. It does not apply to, and must not expand into:
supabase-integration / supabase),frontend-design / impeccable).Higgsfield is the asset engine, not the app builder. Stay in your lane: pictures, motion, and sound.
The plugin verifies Higgsfield is connected and signed in before any work starts (the
using-designer-workflow hard gate). Do not re-run the full gate here — just generate. If a generation
call returns an auth/connection error mid-build, surface the same /mcp → higgsfield → Authenticate
sign-in instruction and pause.
Higgsfield MCP tools are deferred — load them with ToolSearch (select:mcp__...higgsfield...) the
first time you need them. Then choose by asset type:
| Need | Higgsfield tool |
|---|---|
| Image, icon, logo, illustration, hero, background, avatar, texture | generate_image |
| Video / motion clip | generate_video |
| Sound effect, music bed, or voiceover | generate_audio |
| 3D asset | generate_3d |
| Sharpen / enlarge an existing asset | upscale_image, upscale_video |
| Transparent overlay (cut out the subject) | remove_background |
| Extend or re-aspect an image for a layout | outpaint_image, reframe |
Helpers: presets_show / models_explore to choose a brand-locked preset/model before generating
so assets stay on-brand and consistent across the app. When the user provides their own local
photo/logo to work from, call media_upload_widget; for a web URL, call media_import_url first and
pass the returned media_id (never paste a raw URL into a generation param).
presets_show) so the set is brand-consistent, then generate each asset at
the right dimensions/aspect for where it lands (hero vs. icon vs. avatar).verify-in-browser applies:
confirm they actually render at mobile + desktop and nothing is broken/blurry/missing.Translate — never surface tool names, prompts, or asset IDs in chat:
"I generated the artwork with our brand presets — a hero image, the nav icons, and a little empty-state illustration — and placed them in. They render cleanly on phone and desktop."
The raw prompts/preset choices, if worth noting, go in the PR, not the chat.
npx claudepluginhub potenlab/designer-workflow --plugin designer-workflowGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.