R2 Prompt Loader

R2 Prompt Loader — ComfyUI Custom Node

ComfyUI custom node for reading/writing text files and browsing/uploading images from/to Cloudflare R2 (S3-compatible) object storage, with a built-in thumbnail gallery.

Features

• Text operations: Browse directories, read .txt files, write/save text prompts to R2
• Image gallery: Thumbnail gallery with lazy loading, batch fetching (8 per batch), scroll-to-load
• Image operations: Browse, delete, and upload images from/to R2
• S3-compatible: Works with any S3-compatible object storage (Cloudflare R2, AWS S3, MinIO, etc.)

Installation

Clone into your ComfyUI/custom_nodes/ directory:

cd ComfyUI/custom_nodes/
git clone https://github.com/<YOUR_USERNAME>/r2-prompt-loader
pip install -r r2-prompt-loader/requirements.txt

Restart ComfyUI.

Usage

1. Add R2 Prompt Loader node (category: R2 Tools)
2. Enter your S3 endpoint, access key, secret key, and bucket name
3. Click 登录 to test the connection
4. Browse directories and files, or use the image gallery

Node Inputs

| Input | Type | Description | |-------|------|-------------| | endpoint | STRING | S3 endpoint URL | | access_key | STRING | Access Key ID | | secret_key | STRING | Secret Access Key | | bucket | COMBO | Bucket name | | directory | COMBO | Directory prefix (click ↻ to list) | | file | COMBO | .txt file selection (click ↻ to list) | | image_file | COMBO | Image file selection (click to preview) | | text_to_save | STRING | Text content to write to R2 | | save_filename | STRING | Save path (default: prompts/new_prompt.txt) | | upload_image | IMAGE | Connect an image node output, click 存图 to upload |

Node Outputs

| Output | Type | Description | |--------|------|-------------| | text | STRING | File content or upload result message | | image | IMAGE | Selected image tensor [1, H, W, 3] (float32 0-1) |

Buttons

• 登录 — Test R2 connection, auto-refresh lists on success
• 写入 — Save text to R2
• 删除 — Delete selected .txt file
• ↻ Dir / ↻ Files — Refresh directory/file lists
• 删图 (red) — Delete selected image
• 存图 (blue) — Queue workflow to upload connected image to R2

File Structure

custom_nodes/r2-prompt-loader/
├── __init__.py          # ComfyUI extension entry point
├── node.py              # R2PromptLoader node definition
├── server.py            # Backend API routes (9 endpoints)
├── requirements.txt     # Python dependencies
└── js/
    └── main.js          # Frontend JS extension

API Routes

Mounted at /r2-prompt-loader/api/:

| Route | Method | Description | |-------|--------|-------------| | /test | POST | Test R2 connection | | /list_dirs | POST | List subdirectories | | /list_txt | POST | List .txt files | | /list_images | POST | List image files | | /get_image | POST | Get full-resolution image | | /get_thumbnail | POST | Get 200px JPEG thumbnail (~10KB) | | /write_txt | POST | Write text file | | /delete_txt | POST | Delete text file | | /delete_image | POST | Delete image file |

License

MIT