Prompt Styler Fantasy-Setting

Type a character, pick a world, get the world. This node's dropdown is a list of fantasy places - Atlantis, Enchanted Forest, Fairyland, Middle Earth, Mordor, Narnia, Neverland, Shambhala, and 38 more - and it wraps your subject in the scenery, lighting, and mood of whichever one you choose. wolfden's own README uses "Fantasy Setting – Neverland" as a showcase example, so this is one of the pack's marquee nodes.

It's the fast way to answer "put this figure somewhere magical" without writing three sentences of environment yourself. You handle the who; the style handles the where.

How it works

Every setting is a template row in a JSON file: a prompt field with a {prompt} placeholder plus a negative_prompt. The node substitutes your positive text into the placeholder and returns the combined string. Leave text_negative empty and the template's negative is used; fill it and yours is merged in. It's pure text templating - no model call, no download, no key. The picture itself is made downstream by your checkpoint; the node just writes the prompt.

Worth knowing: because these are named, well-known settings, SDXL often already has strong associations for them from its training data. "Mordor" pulls volcanic desolation, "Enchanted Forest" pulls dappled light and glowing flora. The template reinforces and steers those associations rather than inventing them from scratch.

The inputs and outputs that matter

• style - the world. The node's entire purpose; 46 to pick from.
• text_positive - your subject. "a lone knight," "a small fox," "a sailing ship." The setting supplies everything around it.
• text_negative - optional. Blank uses the template's default; add your own to knock out anything specific you keep getting.

log_prompt (No/Yes) prints the finished prompt to the console - handy for seeing how a setting like "Shambhala" got expanded.

The node outputs positive_prompt_text_g and negative_prompt_text_g. Feed the positive into your positive CLIP Text Encode and the negative into the negative one, then sample as usual.

Installing it

ComfyUI Manager: search SDXL Prompt Styler (the wolfden fork), install, restart.

Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/wolfden/ComfyUi_PromptStylers

Restart ComfyUI. Nothing to pip install, no model files to fetch. After the restart the styler nodes show up under a Style Prompts submenu.

Common issues

The first option is a passthrough. none byFantasy-Settings applies no styling. If your generation looks unstyled, make sure you actually picked a world.

Trademarked worlds, generic results. Names like Narnia and Middle Earth steer toward the vibe, not a licensed reproduction - SDXL renders "a snowy lamppost in a wood," not the film. That's usually what you want, but don't expect frame-accurate movie scenes.

Mismatched subject and setting. A neon-lit robot in "Fairytale" fights itself. Pair subjects that belong in the world, or expect the model to average two ideas that don't blend.

SDXL-family only, really. These templates are SDXL-era and count on tag-style prompting and a live negative prompt. On Flux or Z-Image the negative is inert at CFG 1 and the descriptors read as plain instruction, so the setting comes through weakly. Keep it on SDXL checkpoints.