stable-diffusion-helper

Name: stable-diffusion-helper
Availability: InStock
Author: Jack

Configure Stable Diffusion models, prompts, and ADetailer for photorealism.

16|7|Updated Nov 20, 2025

Search Tags:#prompt-engineering #VRAM #photorealism #stable-diffusion #ADetailer

Ready to run

npx skills add https://github.com/jackspace/ClaudeSkillz --skill stable-diffusion-helper

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: stable-diffusion-helper
Source: https://github.com/jackspace/ClaudeSkillz/tree/main/skills/stable-diffusion-helper
Command: npx skills add https://github.com/jackspace/ClaudeSkillz --skill stable-diffusion-helper

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides expert advice on model selection, prompts, and configurations to achieve photorealistic SD results and avoid common issues.

Core Features & Use Cases

Model selection: Compare SD models and pick the best fit.
Prompt engineering: Formulas for realism and anti-cartoon prompts.
Performance tips: VRAM optimization, batch sizing, and prompts tuning.
Troubleshooting: Fix eyes, skin, hands, etc.

Quick Start

Ask for model recommendations and a ready-to-use prompt template for a portrait or product shot.

Frequently Asked Questions about stable-diffusion-helper

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I get photorealistic results from Stable Diffusion?▼

Photorealistic Stable Diffusion results depend on three factors: selecting a model trained for realism (not general-purpose), engineering prompts with anti-cartoon keywords and specific detail descriptors, and tuning CFG scale to 4.5 for balance between prompt adherence and natural variation. ADetailer passes refine faces and hands.

What's the best CFG scale and model settings for portrait photography with Stable Diffusion?▼

Use CFG scale 4.5 for photorealistic portraits to avoid overconstraint artifacts. Pair this with a photorealism-focused model, reduce batch size if VRAM is limited, and enable ADetailer with separate face and hand passes to correct common distortions in eyes, skin texture, and fingers.

How do I optimize Stable Diffusion for VRAM constraints?▼

VRAM optimization for Stable Diffusion involves reducing batch size, selecting lighter models, and tuning prompt length. Performance tips include disabling unnecessary features during iteration, using memory-efficient schedulers, and staging ADetailer passes only after base generation succeeds.

Why do hands and faces look wrong in my Stable Diffusion images?▼

Hand and face distortions are common in Stable Diffusion because the base model underfits fine details. ADetailer solves this by running dedicated passes over face and hand regions with specialized prompts and higher detail guidance after base generation, then compositing corrected regions back.

Can I use prompt engineering formulas to avoid cartoon-like output in Stable Diffusion?▼

Yes. Anti-cartoon prompt formulas combine negative keywords (avoid stylization terms), positive detail descriptors (film grain, skin texture, lighting specifics), and model selection. Pairing these with CFG 4.5 and ADetailer refinement produces photorealistic portraits and product shots consistently.

What should I do if Stable Diffusion troubleshooting doesn't fix quality issues?▼

Systematic troubleshooting checks model fit for your use case, prompt formula alignment with realism goals, VRAM availability for ADetailer passes, and CFG scale calibration. If base generation fails these checks, reconfigure incrementally: swap model, refine anti-cartoon keywords, enable batch-size reduction, then add ADetailer.