This repo packages Bria AI skills that let your agent generate, edit, remove backgrounds from, and restyle images with direct model control. The skills cover both simple prompts and structured VGL JSON so you can steer objects, lighting, composition, and pipelines more precisely.
Builders who use Claude Code, Cursor, or Codex and want image generation and editing built into their agent workflow.
You can create and transform images from inside your agent instead of switching to a separate image tool.
The `bria-ai` skill handles text-to-image, object edits, masking, relighting, restyling, upscaling, and product photography.
The `remove-background` skill uses RMBG 2.0 for transparent PNGs, cutouts, and foreground extraction.
The `vgl` skill uses structured JSON to define objects, lighting, camera, composition, and style for reproducible image generation.
The `ad-delayer` skill breaks flat ads into editable layers like background, product, logo, headline, copy, and CTA.
The `image-utils` skill provides resize, crop, composite, watermark, and format conversion workflows with Pillow.
/plugin marketplace add Bria-AI/bria-skill /plugin install bria-ai@bria-skills
npx skills add bria-ai/bria-skill
export BRIA_API_KEY="your-key-here"
claude mcp add bria-skills -- npx -y bria-skills
Generate, edit, and precisely control images directly from your AI coding agent using Bria.ai's commercially-safe models. Unlike black-box generators, Bria gives you fine-grained control: edit specific objects, mask regions, control lighting/season/style independently, and chain operations in pipelines.
Works with: Claude Code, Cursor, Cline, Codex, and 37+ other agents
Claude Code (plugin): add the marketplace, then install the plugin:
/plugin marketplace add Bria-AI/bria-skill
/plugin install bria-ai@bria-skills
The marketplace also bundles the remove-background, ad-delayer, vgl, and image-utils plugins — install any of them the same way (e.g. /plugin install remove-background@bria-skills).
Other agents (Cursor, Cline, Codex, and 37+ more):
npx skills add bria-ai/bria-skill
Sign up at platform.bria.ai and create an API key.
export BRIA_API_KEY="your-key-here"
Or just start using the skill — it will walk you through setup automatically.
Ask your agent to generate images, remove backgrounds, edit photos, and more. The skill handles the rest.
| Skill | Description |
|---|---|
| bria-ai | Controllable image generation & editing — 18+ capabilities including text-to-image, object-level editing, background removal, upscaling, restyling, relighting, and product photography |
| remove-background | Dedicated background removal — transparent PNGs, cutouts, and foreground extraction powered by RMBG 2.0. Fastest path for background removal tasks |
| vgl | Maximum generation control — structured JSON (Visual Generation Language) that explicitly defines objects, lighting, camera, composition, and style for deterministic, reproducible results |
| ad-delayer | Flat ads back into editable layers — background, product shot, logo, headline, copy, and CTA recovered as separate assets with position, size, and typography |
| image-utils | Classic image manipulation — resize, crop, composite, watermarks, format conversion with Python Pillow |
| Capability | Control Level | Example |
|---|---|---|
| Generate images from text | Prompt + aspect ratio + VGL structured JSON | "A modern office with natural lighting, 16:9" |
| Edit specific objects | Object-level — add, replace, or remove by name | "Replace the red mug with a blue one" |
| Edit by instruction | Image-level — natural language control | "Change the mug color to red" |
| Edit masked regions | Pixel-level — paint exactly where to change | Inpaint only the sky, leave everything else |
| Remove backgrounds | Automatic — clean transparent PNGs | One API call, no manual masking |
| Control lighting | Independent — change light without touching content | Golden hour, spotlight, dramatic shadows |
| Control style | Independent — restyle without changing structure | Oil painting, anime, cartoon, 3D render |
| Control season | Independent — change season/weather only | Spring blossoms, winter snow |
| Expand images | Directional — extend boundaries with context | Outpaint to 16:9 from a square |
| Product photography | Scene control — place products in any setting | "Modern kitchen countertop, morning light" |
| Upscale resolution | 2x or 4x — preserves details | Super resolution enhancement |
| Pipeline chaining | Multi-step — compose operations in sequence | Generate -> remove BG -> lifestyle shot |
Building a website? Ask: "Generate a hero image for a SaaS landing page, 16:9, modern and clean"
E-commerce catalog? Ask: "Remove the background from this product photo and place it in a kitchen scene"
Batch assets? Ask: "Generate product photos for these 5 items and create transparent PNGs for each"
Presentations? Ask: "Create a slide visual showing data analytics, blue gradient, corporate style"
claude mcp add bria-skills -- npx -y bria-skills
Download the skill folders to ~/.cursor/skills/. Each skill is a folder containing a SKILL.md file.
Copy the skills/ directory contents to your agent's skills directory.
pip install Pillow requestsMIT
Sign in to join the discussion.
No comments yet. Be the first to say what this is good for.
Topic in, narrated explainer video out. A Claude Code / Codex skill that turns any topic into a black-canvas motion-graphics explainer video with TTS voiceover, subtitles and a chapter progress bar. Chinese or English; every frame drawn in code with Remotion.
Public repository for Agent Skills
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
Taste-Skill - gives your AI good taste. stops the AI from generating boring, generic slop
Agent skill for beautiful, verifiable architecture, workflow, sequence, data-flow, and lifecycle diagrams—self-contained HTML with motion and crisp export.
Rebuild the object in a reference image as a code-only, procedural, quality-gated, animation-ready Three.js model. Token-efficient image-to-3D.