Anima Prompt Studio Advanced

Anima's whole pitch is that tags and natural language coexist in one prompt - and that means your prompt has structure: quality tags up front, artists, triggers, a body, trailing style. Anima Prompt Studio Advanced is the editor that keeps that structure explicit instead of one giant text box. It's the "big" editor of the pack: reorderable positive and negative fields, NAIA fill, resolution control, wildcards, and Mod Guidance routing, all feeding ten string outputs. If you've seen a version of this with one socket instead, that's v2 - this is the original, and it's still the right tool for plenty of workflows.

What it is

You edit prompts as separate field groups - quality, artist, trigger, general, and a NAIA field on both the positive and negative sides. Fields can be added, removed, reordered, enabled, or disabled, and the node remembers their order in the workflow. The trigger field can take connected trigger_words from a LoRA preset and pin to the front. Typing __ opens wildcard autocomplete, and the editor highlights quality tags, artists, characters, and unknown tags in different colors.

The key difference from the plain Prompt Studio: it outputs the full set of strings a serious workflow needs - positive_prompt, negative_prompt, both Mod Guidance quality outputs, both Mod Guidance booleans, metadata_prompt, metadata_negative_prompt, and width/height for the latent. Ten sockets, each with a job.

The inputs worth knowing

Most required inputs are internal JSON (the advanced_fields payload, resolution buckets, wildcard seed) that the front-end manages. The ones with user meaning:

• use_naia / consume_naia_on_queue - the NAIA fill machinery. The front-end turns use_naia into the "Fill from NAIA" button; consume_naia_on_queue is one-shot mode so loaded workflows reuse the stored prompt instead of re-calling NAIA.
• use_anima_mod_guidance / use_negative_anima_mod_guidance - when true, the quality fields are pulled out of the main prompt and routed to the Mod Guidance outputs, to be applied as a model patch downstream.
• resolution_bucket / resolution_size - the resolution picker, with buckets from 512 up to 1536 plus Custom; saved workflows store the resolved size so results reproduce.
• wildcard_mode / wildcard_seed / wildcard_seed_after_generate - wildcard expansion control. General mode expands deterministically; Sequential uses seed % option_count.

How it fits a workflow

The outputs wire into the classic ComfyUI path directly: positive_prompt and negative_prompt into CLIP text encode (or into Anima Prompt Data Conditioning via a v2 dict), width/height into Empty Latent Image, and the Mod Guidance strings into a Spectrum Mod Guidance node. This is the "I want the editor but I'm building my own graph" option - v2 exists for the all-in-one dict flow.

Install and gotchas

It's one node in the pack:

cd ComfyUI/custom_nodes
git clone https://github.com/n0va39/ComfyUI-EasyUseAnima
cd ComfyUI-EasyUseAnima && pip install -r requirements.txt

Restart ComfyUI (Manager: "ComfyUI EasyUse Anima").

Gotchas: pin_trigger_tags_to_front is a legacy flag - the per-field Pin buttons on trigger fields are what actually control placement now, so don't chase it. NAIA fill needs a local NAIA service on 127.0.0.1 (the default) or the global "Allow remote API" setting switched on; the node itself won't reach a remote host by default. And if the ten outputs feel like overkill, that's fine - that's literally why v2 and the unpack node exist.