ComfyUI-d2-send-eagle

ComfyUI-d2-send-eagle
★ 19

图像管理Eagle图片导出PNG/WebP
将 ComfyUI 生成的图片发送到 Eagle，并把正/负向提示文本记录为 Eagle 备忘，支持 PNG 与 WebP 格式，方便集中管理与归档。
💡 将生成图片与提示文本一键同步到 Eagle 管理库。
🍴 8 Forks💻 Python🔄 2025-03-10
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/8b0992d318c3
📦 requirements.txt
piexif
📄 README

D2 Send Eagle

[English][日本語][繁体中文]

A custom node for sending images generated in ComfyUI to the image management software Eagle.

Please also try D2 Nodes ComfyUI, which contains a collection of useful custom nodes.

Sample Workflows

Basic Configuration Workflow

Workflow Using D2 Nodes ComfyUI

  • Gets checkpoint name from D2 Checkpoint Loader and removes the “.safetensors” extension using D2 Filename Template
  • Retrieves generation parameters from d2_pipe

    • Main Features

  • Sends images received through image to Eagle
  • Records text received through positive and negative as Eagle memos
  • Supports both png and webp formats

    • Generation Parameters Recorded as Image and Eagle Comments

  • Parameters such as positive, negative, CFG, steps are retrieved from the KSampler connected to D2 Send Eagle
  • May not be retrievable depending on the workflow
  • Parameters can also be specified through d2_pipe

    • Supported KSamplers

    The following KSampler types are supported.

    You can add more by editing config.yaml.

  • KSampler
  • KSamplerAdvanced
  • KSampler With Refiner (Fooocus)
  • BNK_TiledKSampler
  • KSampler (Efficient)
  • GenerateNAID

    • Differences Between png and webp Formats

  • png format
  • Can save ComfyUI workflows
  • Cannot display in StableDiffusion webui A1111’s PNGInfo
  • webp format
  • Cannot save ComfyUI workflows
  • Can display in StableDiffusion webui A1111’s PNGInfo

    • Installation

    Using ComfyUI Manager

  • Open ComfyUI Manager
  • Click Custom Nodes Manager
  • Search for D2 Send Eagle
  • Click Install
  • Restart ComfyUI

    • Using Command Prompt

  • Open Command Prompt
  • Navigate to {ComfyUI installation folder}/custom_nodes
  • git clone https://github.com/da2el-ai/ComfyUI-d2-send-eagle

    • Inputs / Options

  • images
  • Images to save
  • d2_pipe
  • Collection of generation parameters
  • positive
  • Positive prompt
  • negative
  • Negative prompt
  • format
  • Choose from webp / png / jpeg formats
  • lossless_webp
  • Choose between lossless and lossy. Active when webp is selected.
  • compression
  • Specify compression rate. Active when selecting lossy for webp or when using jpeg.
  • save_tags
  • Choose whether to save as Eagle tags
  • None: Don’t save
  • Prompt + Checkpoint: Save prompts and model name
  • Prompt: Save prompts
  • Checkpoint: Save model name
  • filename_template
  • Specify filename format
  • Default is {model}-{width}-{height}-{seed}
  • Available parameters: width, height, model_name, steps, seed
  • eagle_folder
  • Specify Eagle folder name or folder ID. Creates new folder if it doesn’t exist

    • Other Features

    Local Image Saving

    In addition to sending to Eagle, images are also saved locally in the following folder.

    This folder name cannot be changed.

    ./ComfyUI/output/YYYY-MM-DD/YYYYMMDD_HHMMss_SSSSSS-{FinalImage_width}-{FinalImage_height}.webp

    Change Log

  • 2024/12/28
  • Added support for d2_pipe in d2-node-comfyui
  • 2024/11/13
  • Added JPEG format support
  • 2024/10/04
  • Added toggle button to hide image preview
  • Because enlarging preview also enlarged memo_text which was inconvenient
  • Added IMAGE to output
  • Just a passthrough. Not necessary but added anyway
  • 2024/09/29
  • Added support for NovelAI generation node GenerateNAID
  • Made settings configurable through config.yaml
  • Added Exif prompt information recording for webp format
  • 2024/08/19
  • Fixed issue with model name not being retrieved when using Unet Loader
  • 2024/08/04
  • Initial release

    • Acknowledgements

    D2 Send Eagle is an extension of an existing custom node ComfyUI-send-eagle-slim, modified to my preferences. Thanks to Shingo.T for creating such an excellent custom node.