Bria AI Skill — All Agent Compatible

Generate and edit images with your agent.

One skill. Twenty capabilities. Generate product shots, remove backgrounds, restyle scenes, upscale to 4K — just describe what you need. No prompts to engineer. No API docs to read.

npx skills add bria-ai/bria-skill

Download the Skill

20+

Image capabilities

One

Natural language instruction

Zero

Manual API calls needed

How it works

The AI handles the complexity. You handle the vision.

Bria's skill wires the model directly into a professional image generation and editing API. The LLM translates your intent into precise API calls — invisibly.

1

You describe what you want

Speak naturally. "Make this product photo look like it's on a marble kitchen counter, then remove the background." No technical parameters. No format juggling.

plain language input

2

The LLM maps intent to capability

Claude understands context, picks the right Bria endpoint — RMBG-2.0, FIBO, Edit — chains operations if needed, and handles error recovery autonomously.

autonomous orchestration

3

You get production-ready images

Commercially safe, royalty-free output — ready for your campaign, e-commerce PDP, or design pipeline. Download, iterate, or chain into the next step.

commercially safe output

Real conversation, real output

> Generate a high-fashion editorial shot. Avant-garde white blazer, minimalist concrete curves, starburst sun effect.

agent: Generating with FIBO model. Scene: architectural concrete backdrop, dramatic starburst lighting, fashion editorial framing...

✓ Image generated successfully.

output image — Generated with FIBO — commercially safe, production ready.

> Now create a cutout of the model — transparent background, clean edges.

agent: Running RMBG-2.0 on the generated image. Detecting edges, separating foreground...

✓ Background removed. Transparent PNG ready.

Every visual job, covered

The agent picks the right tool automatically. You never choose an endpoint — you just describe the outcome.

Background removal

Instant transparent PNG cutouts via RMBG-2.0. Works on products, people, and complex edges.

Text-to-image generation

FIBO model generates hero images, banners, illustrations, and ad creatives from a description.

Image editing by instruction

Change colors, swap objects, alter scenes — no mask, no Photoshop. Just describe the edit.

Lifestyle product shots

Drop a product into a contextual scene — kitchen, outdoor, studio — with natural lighting and perspective.

Outpainting & expand

Extend image boundaries to any aspect ratio. Make a 1:1 into a 16:9 with seamless AI fill.

Upscale & enhance

2x or 4x super-resolution for print-ready quality. Also restores and sharpens older images.

Restyle & reseason

Transform the visual mood — oil painting, anime, 3D render — or swap seasons from summer to winter.

Batch pipelines

Chain multiple operations across multiple images — generate, edit, remove background, deliver — all in one instruction.

“You're not calling an API. You're giving your agent a full creative studio — generate, edit, and transform images the same way it handles any other task: autonomously, in context, on demand.”

Bria AI Skills — Agent-native image intelligence

Installation Steps

Download the skill file

Download the bria-ai.skill file from the Bria website. Save it somewhere easy to find, like your Downloads folder.

Open Claude

Launch the Claude desktop app. Start a new conversation or open an existing one.

Drag the file into the chat

Drag and drop the bria-ai.skill file directly into the Claude chat window, just like you would attach any file. Alternatively, click the attachment (paperclip) icon and select the file from your Downloads folder.

Accept the installation

Claude will show a preview of the skill with its name, description, and the tools it includes. Click "Install" (or "Accept") to confirm. The skill is now active in your Claude session.

Use the skill

Now you can use the skill to create something amazing with Bria AI.