AWS Nova Canvas MCP Server
<h2 align="center">
AWS Nova Canvas MCP Server
</h2>
<div align="center">
<img src="https://img.shields.io/badge/Python-3.12-3776AB?logo=python"/>
<img src="https://img.shields.io/badge/Amazon-Bedrock-FF9900?logo=amazon&logoColor=white"/>
</div>
An MCP server that allows you to generate and edit images using the Nova Canvas model of Amazon Bedrock.
## Features
- Text to Image
- Image Inpainting
- Image Outpainting
- Image Variation
- Image Conditioning
- Color Guided Generation
- Background Removal
- Show Image Thumbnails
## Installation
### Claude Desktop Setup
1. Configure Claude Desktop
* Click on **Claude > Settings** from the Claude Desktop menu.
* When the popup appears, select **Developer** from the left menu, and click the **Edit Settings** button.
* This will open a folder containing the settings file. The name of this settings file is:
* `claude_desktop_config.json`
<div align="center">
<img src="https://blog.kakaocdn.net/dn/bIl5q9/btsM3U5Vjw5/aGruWqP3wNmWZ1sKrnhbPk/img.png" width="70%">
</div>
3. Add the following content to the settings file (Python version):
- python version
```json
"nova-canvas": {
"command": "uvx",
"args": [
"aws-nova-canvas-mcp"
],
"env": {
"AWS_PROFILE": "YOUR_AWS_PROFILE"
}
}
```
> ✅ Only AWS_PROFILE is required. Other variables like AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_REGION, and PORT are optional and not necessary if your AWS profile is set correctly.
>
> ⚙️ If the setup is completed successfully, you can see that the "nova-canvas" item has been added in **Claude > Settings > Developer tab**.
> ⚠️ **Important:** MCP settings only work on the **Claude desktop app, not the Claude web browser version**
## Image Save Location
By default, all generated or edited images will be saved in the following directory:
* **macOS / Linux**: `~/Desktop/aws-nova-canvas`
* **Windows**: `C:\Users\YourUsername\Desktop\aws-nova-canvas`
> 📁 If no image save path is specified, the application will automatically create and use the folder above.
<div align="center">
<img src="https://blog.kakaocdn.net/dn/bpUWLj/btsM4kJZC6v/HHQfQctKsevWnK6LCKEkv0/img.png" width="70%">
</div>
## Usage Example
<div align="center">
<img src="https://blog.kakaocdn.net/dn/uNi8L/btsM4pEjswV/hSfxo1gHzPvpXPsEEyuijk/img.gif" width="70%">
</div>
## Limitations
- Prompt text supports up to 1024 characters
- Image generation allows up to 3 images at a time
- Image variation requires 1-5 reference images
- Color guide supports 1-10 color codes
## License
MIT License
TDQS
Scored across 8 tools
Each tool has a clearly distinct purpose within the image processing/generation domain. Background removal, color-guided generation, image conditioning, image variation, inpainting, outpainting, show_image, and text_to_image all target specific, non-overlapping operations. The descriptions clearly differentiate their functions, making tool selection unambiguous.
All tool names follow a consistent snake_case pattern with clear verb_noun or noun_verb structures (e.g., background_removal, color_guided_generation, show_image). The naming is predictable and readable throughout the set, with no mixing of conventions or stylistic deviations.
With 8 tools, the count is well-scoped for an image processing/generation server. Each tool serves a distinct and valuable function, covering core operations like generation, editing, and viewing without redundancy. The number is neither too sparse nor overwhelming for the domain.
The tool set provides comprehensive coverage for image generation and manipulation, including text-to-image, variations, conditioning, inpainting/outpainting, background removal, and viewing. A minor gap exists in lacking explicit editing tools like cropping or filtering, but core workflows are well-supported, and agents can likely work around this.