Halloween Styler

October's coming and you want "haunted house," not "a house that is haunted-ish." HalloweenStyler puts 51 spooky presets on one dropdown - Classic Haunted House, Witch's Cauldron, Ghostly Apparitions, Zombie Apocalypse, Frankenstein's Lab, Mummy's Tomb, Possessed Doll, Haunted Asylum, Grim Carnival, Voodoo Doll. Type a subject, pick a flavor of horror, and the node writes the prompt and negative for you. It's one of the per-theme stylers in aegis72/comfyui-styles-all, the pack that preserved azazeal04's removed SDXL Prompt Styler styles.

What it actually is

Prompt styler, pure and simple. Each Halloween option is a JSON template with a {prompt} placeholder plus a negative prompt. The node drops your subject into the placeholder and appends the style's negative to yours. No model logic, no API - text in, two strings out.

The menu's best feature is breadth: it covers the whole spooky spectrum rather than one mood. Pumpkin Patch and Zombie Wedding for the light end; Demon's Inferno, Cursed Mirror, and Vampire's Feast for the heavy end; plus genre odds like Horror Film Director, Spectral Opera House, and Pirate's Curse that you'd never think to prompt by hand. Handy for a themed batch - greeting cards one render, grim dark the next.

The inputs that matter

• text_positive - your subject; the horror scene wraps around it.
• Halloween - the 51-option dropdown; None passes through.
• log_prompt - on by default; prints before/after to the console.

Outputs: text_positive (styled) and text_negative (your negative plus the style's, comma-joined). Both STRING - straight into CLIPTextEncode.

Installing it

ComfyUI Manager (search ComfyUI-styles-all), or:

cd ComfyUI/custom_nodes
git clone https://github.com/aegis72/comfyui-styles-all

Then restart ComfyUI. No requirements.txt, no models.

Where people get burned

• Negatives stack. Each preset ships its own negative, so keep yours lean or the combined string gets heavy.
• Restart after edits. The dropdown is built from data/Halloween/Halloween.json at startup.
• SDXL-era keywords. Strong on SDXL-lineage checkpoints, weak on Flux and other LLM-encoded models - same story as every node in this pack.