Repolode
Submit
Arekejoker

Arekejoker/qwen21-nsfw-pose-skill

Skill-based recipe kit for generating NSFW and pose-controlled images with Qwen-Image 2.1 in ComfyUI, with a 136-pose 3D mannequin library.

View on GitHub
Stars
2,624
Forks
504
Open issues
4
Last commit
2 days ago

What is it?

qwen21-nsfw-pose is a set of prompt structures, LoRA settings, workflows and helper scripts for generating adult-oriented images locally with the Qwen-Image 2.1 int8 model in ComfyUI. It packages a prompt template with required elements, a default NSFW LoRA at strength 1.0, and sampler settings (er_sde, beta scheduler, CFG 2.5, 50 steps, 1152x1728 portrait).

The project offers three routes: text-to-image with an NSFW LoRA, image editing that keeps an existing character's face and clothing while changing pose, and precise pose control using a 136-pose 3D mannequin library through VNCCS nodes. A doctor.py script checks ComfyUI connectivity, models, LoRAs, nodes and pose library before generation.

It is written as an agent skill for Claude Code, Codex, Cursor and OpenClaw, and can also be followed manually. Setup instructions, hardware notes and sample metadata are included. Model and LoRA licenses belong to their original authors.

Key features

  • 01Qwen-Image 2.1 int8 workflow for ComfyUI with text-to-image and edit modes
  • 02Default NSFW LoRA with recommended sampler and step settings
  • 03136 3D pose presets with a pose-selection guide from mild to bold
  • 04VNCCS mannequin control for pose changes while keeping identity
  • 05doctor.py self-check for models, LoRAs, nodes and pose library

Use cases

  • Generating new figure images from text prompts on a local GPU
  • Changing the pose of an existing character image while preserving face and outfit
  • Reproducing a specific pose from a 3D mannequin library in a scene prompt
  • Using an AI coding agent to run the image-generation workflow as a skill

Topics

#image-generation#comfyui#qwen-image#pose-control#agent-skill#lora#stable-diffusion#openclaw

FAQ

What is qwen21-nsfw-pose-skill?
Skill-based recipe kit for generating NSFW and pose-controlled images with Qwen-Image 2.1 in ComfyUI, with a 136-pose 3D mannequin library. qwen21-nsfw-pose is a set of prompt structures, LoRA settings, workflows and helper scripts for generating adult-oriented images locally with the Qwen-Image 2.1 int8 model in ComfyUI. It packages a prompt template with required elements, a default NSFW LoRA at strength 1.0, and sampler settings (er_sde, beta scheduler, CFG 2.5, 50 steps, 1152x1728 portrait).
Is qwen21-nsfw-pose-skill free to use? What license does it have?
qwen21-nsfw-pose-skill is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is qwen21-nsfw-pose-skill still maintained?
Actively maintained. The last commit to qwen21-nsfw-pose-skill was on Oct 17, 2026.
What can I use qwen21-nsfw-pose-skill for?
Generating new figure images from text prompts on a local GPU Changing the pose of an existing character image while preserving face and outfit Reproducing a specific pose from a 3D mannequin library in a scene prompt Using an AI coding agent to run the image-generation workflow as a skill
Where can I get qwen21-nsfw-pose-skill?
The source code and installation instructions are on GitHub: https://github.com/Arekejoker/qwen21-nsfw-pose-skill

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects