Node packVerifiedExternal

ComfyUI-RookieUI

The ultimate A1111-style sidebar. Bring the classic UX you know and love into ComfyUI.

Sourcegithub.com/rookiestar28/ComfyUI-RookieUI

Maintainer@rookiestar28Bgrade

Package profile

Status: Verified
Publisher: @rookiestar28
Source: GitHub
Stars: 95

README

ComfyUI-RookieUI

ComfyUI-RookieUI is a ComfyUI custom node extension that reproduces an A1111-style sidebar workflow while keeping inference inside native ComfyUI execution. The project target is not only visual similarity. RookieUI aims to reproduce A1111-style workflow semantics for Stable Diffusion in a ComfyUI host:

prompt / negative prompt parsing and conditioning behavior
sampler / scheduler / seed / CFG mapping
img2img, inpaint, and extras postprocessing flows
integrated ControlNet and ADetailer behavior
PNG metadata embedding, round-trip, and apply workflow
Prompt Workbench authoring tools and XYZ Plot sweep sessions
queue / progress / result UX and host-integrated workflow behavior

The core objective of this project is not merely to replicate the classic UI/UX, but to faithfully reproduce A1111's unique prompt parsing capabilities and image generation characteristics for the Stable Diffusion model family to the greatest extent possible. Even so, RookieUI supports more than just the Stable Diffusion family.

Translation and AI assist now accept only the official OpenAI and MyMemory HTTPS endpoints.
Custom endpoint settings and imports are rejected; provider requests block redirects and unsafe network destinations.
Provider connections retain TLS certificate verification and bypass environment proxy routing to keep API credentials bound to the approved service.
Existing custom endpoint settings require an explicit switch to the official endpoint before execution. API keys remain masked in exports and browser responses.

Fixed default Stable Diffusion prompt encoding failures on newer ComfyUI hosts.
Extras now preserves source transparency through resizing, upscaling, blending, face restoration, and color correction.
Dedicated Qwen Image 2.1 prompt enhancement models are excluded from generation encoder selection, preventing incompatible automatic or manual pairings.
Expanded source compatibility checks for newer ComfyUI, frontend, and workflow-template versions while keeping the documented host compatibility scopes separate.
Updated frontend dependencies to address known security advisories.
Verified saved SD1.5 PNG prompt metadata with ordinary, weighted, multiline, Unicode, and long prompts.

Added official-template-backed Qwen-Image 2.1 txt2img and Qwen-Image 2.1 Edit profiles with a dedicated Qwen Image 2.1 encoder and host-inventory matching for the diffusion model, text encoder, and VAE.
The edit profile accepts one to ten ordered references. Reference 1 is the edit target; users can promote another reference into that slot and choose source-based or custom output sizing.
Added an editable background-removal instruction preset to the same edit flow. RGBA PNG data is preserved through the image-load/save and result-handoff paths when present.
Added scoped source-contract coverage for the Core 0.37 Qwen node contract, bundled frontend 1.53.6, standalone frontend 1.55.12, and workflow templates 0.11.66; this does not replace the separately documented general compatibility baseline.
The Qwen Image 2.1 profiles require the matching diffusion model, Qwen3-VL 8B encoder, and 2.1 VAE. Review the Qwen Research License before using the model.

<summ

Nodes in this pack

20 nodes