Text to Image
Describe what you want and get a generated image — from product visuals to brand posters and creative concepts.
Add AI Image Generator Skill to Your OpenClaw
Plug the AI Image Generator Skill into your OpenClaw agent and unlock image creation, editing, product model images, and background removal — all through natural language.
Copy to your agent to install
Install Topview Skill for me: https://github.com/topviewai/skill.gitQuick Start
ImageDemo
Send this prompt to your agent to install Topview Skill.
Install Topview Skill for me:https://github.com/topviewai/skill.git
AI Image Generator Skill installed successfully
You now have access to these image capabilities through your AI assistant:
Text to Image
AI Image Editing
Background Removal
Supported models
Video models
Seedance 2.0 StyleSora 2Kling 3Veo 3.1Gen-4.5wan2.7Vidu Q3Seedance
Image models
nanobanana2nanobanana proSeedream 5.0Seedream 4.5Seedream 4.0GPT Image 1.5Imagen 4Kontext-ProGrok Image
Supported platforms
OpenClaw
Claude Code
Cursor
Codex
Copilot CLI
Antigravity
Coze
What is AI Image Generator Skill?
AI Image Generator Skill is a capability you can add to your AI assistant through Topview Skill. Once installed, your agent can create images from text, edit existing visuals, remove backgrounds, and generate product model images — all through natural conversation.
What you can do with AI Image Generator Skill
AI Image Editing
Change backgrounds, swap scenes, adjust details, or restyle an image by describing what you want changed.
Product Model Images
Place your product into model-style marketing scenes automatically — great for e-commerce and ads.
Background Removal
Get clean transparent cutouts from any image in one step, ready for downstream workflows.
10+ Image Models
Switch across models like Nano Banana 2, GPT Image 1.5, Imagen 4, and more for the style you need.
Visual Asset Management
All generated images are organized on Topview Board — preview, compare, edit, and share in one place.
Claude Skill protocol
Built on the Claude Skill protocol
This skill package follows the Claude Skill protocol structure, so agents can install, read, and execute it in a predictable way. It is a workflow package rather than an executable installer, which means there is no malware or virus risk in normal use.
View VirusTotal virus scan reportHow to use it
Install Topview Skill for me: https://github.com/topviewai/skill.git
Step 1: Copy the install command and send it to your agent
Install Topview Skill into your agent first so it can call the relevant capabilities.
U
🦞
Step 2: Send your creation request to your agent
Send prompts, images, product photos, portraits, or audio-related requests directly to your agent.
🦞
Step 3: Get the result
Your agent will automatically call the right Topview capability to generate video, image, avatar, or audio results for you.
Frequently asked questions
You can generate images from text descriptions, edit existing images (change backgrounds, swap scenes, adjust details), remove backgrounds for clean cutouts, and place products into model-style marketing scenes.
The skill supports 10+ models including Nano Banana 2, Seedream 5.0, GPT Image 1.5, Imagen 4, Kontext-Pro, and more. Your agent selects the best model for your request, or you can specify one.
Yes. You can describe what you want changed — swap the background, change colors, modify details — and the AI will edit the image precisely based on your instructions.
Upload your product image, and the skill removes the background and places it into a model-style marketing scene. You can browse available avatar templates to find the right fit.
Yes. You can ask your agent to generate multiple visual directions from one idea, or batch-generate a set of brand-consistent assets in a single conversation.
Depending on the model, you can generate images up to 4K resolution with 14 different aspect ratios. The default model Nano Banana 2 supports the widest range of options.
All generated images are organized on Topview Board for preview, comparison, and editing. Your agent will also provide download links after generation.
No. You only need to describe what you want in natural language. The AI handles model selection, parameter tuning, and generation — you just review and iterate.