A collection of utility nodes from the VNCCS project that are useful not only for the project's primary goals but also for everyday ComfyUI workflows.

Example Workflow

An interactive node with a visual widget for controlling camera position. This is the primary node for intuitive angle control. These node is specifically designed for advanced camera control and prompt generation, optimized for multi-angle LoRAs like Qwen-Image-Edit-2511-Multiple-Angles.

• Visual Widget: Allows mouse-based selection of azimuth (rotation around the subject) and distance (rings).
• Elevation Slider: A vertical bar on the right for selecting the elevation angle (-30° to 60°).
• Trigger Word: A square indicator in the bottom-right corner toggles the presence of the trigger in the prompt (green for ON, red for OFF).

Example Workflow

A powerful detailing node that leverages QWEN-Image-Edit2511 model to enhance detected regions (faces, hands, objects). It goes beyond standard detailers by using visual understanding to guide the enhancement process.

• Smart Cropping: Automatically squares crops and handles padding for optimal model input.
• Vision-Guided Enhancement: Uses QWEN-generated instructions or user prompts to guide the detailing.
• Drift Fix: Includes mechanisms to prevent the enhanced region from drifting too far from the original composition.
• Quality of Life: Built-in color matching, Poisson blending (seam fix), and versatile upscaling options.
• Inpainting Mode: specialized mode for mask-based editing or filling black areas.
• Inputs: Requires standard model/clip/vae plus a BBOX_DETECTOR (like YOLO).
• Options: Supports QWEN-Image-Edit2511 specific optimizations (distortion_fix, qwen_2511 mode).

Example Workflow

A robust system for managing and selecting models (LoRAs, Checkpoints) directly within ComfyUI, with support for Civitai and HuggingFace.

This node acts as the backend for the system. It connects to a HuggingFace repository containing a model_updater.json configuration file, which defines the available models and their download sources.

• Repo ID: Specify the HuggingFace repository ID.
• Downloads: Handles downloading models in the background with queue support.
• Civitai Support: Supports API Key authentication for restricted Civitai models.

👉 Configuration Guide: How to create your own model repo

The companion node for selecting models. It provides a rich Graphical User Interface.

• Visual Card UI: Displays the selected model's name, version, installed status, and description in a clean card format.
• Smart Search: Clicking the card opens a modal with a searchable list of all available models in the repository.
• Status Indicators: Shows clear indicators for "Installed", "Update Available", "Missing", or "Downloading".
• One-Click Install/Update: Allows downloading or updating models directly from the list.
• Universal Connection: Outputs a standard relative path string that is fully compatible with standard ComfyUI nodes. You can connect it directly!

👉 Usage Guide: How to use Selector with Standard Loaders

A helper node to simply extract and visualize the crops. Useful when you need extract bbox detected regions but don't want to run whole facedetailer.

1. Copy the ComfyUI_VNCCS_Utils folder into your ComfyUI custom_nodes directory.
2. Restart ComfyUI.

The node outputs a string (Prompt) that should be connected to your CLIP Text Encode or used for concatenation with other prompt parts. It's ideally suited for LoRAs trained on multi-angle datasets.