Anima Prompt Builder

Anima's model card is unusually specific about prompt order: quality/meta tags first, then triggers and artists, then the body, then trailing style - and the community has strong feelings about all of it (score tags narrow the distribution; drop the quality prefix once you name an artist). Anima Prompt Builder is the node that encodes that ordering so you don't hand-assemble it every time. Type each chunk in its own field; it combines them in the ANIMA-correct sequence. No UI tricks, no autocomplete - just deterministic assembly. That's the whole product, and it's a genuinely useful one.

What it does

Five text fields go in:

• quality_tags - the leading quality block
• trigger_and_artist_tags - manual model triggers and @artist tags
• lora_trigger_tags - one-line trigger tags from a LoRA manager
• prompt - the main body (the expected home for NAIA output)
• trailing_quality_tags - trailing style/location tags

Two booleans reshape the result. pin_trigger_tags_to_front (default false) keeps triggers and artists at the very front instead of behind the quality block - useful when a model's trigger only fires if it leads the prompt. use_anima_mod_guidance (default false) is the interesting one: when true, the quality fields are excluded from the assembled prompt and sent out the anima_mod_guidance_quality_tags output instead, to be applied through Spectrum's Anima Mod Guidance model patch. Same idea as the AIO generation flow, but in a plain text-node form.

Outputs

• prompt - the final assembled positive prompt.
• anima_mod_guidance_quality_tags - quality text routed for Mod Guidance (only populated when use_anima_mod_guidance is true).
• use_anima_mod_guidance - the boolean, passed through so a downstream workflow can branch on it.
• metadata_prompt - prompt text for metadata purposes, independent of Mod Guidance routing and metadata filters.

There's no negative prompt here, no width/height, no LoRA stack - if you need those, this isn't the node (see Prompt Studio Advanced or the v2/dict flow).

Why you'd reach for it

This is the unglamorous, dependable middle: you want ANIMA's recommended ordering and the Mod Guidance split, but you don't need a fancy editor. Wire prompt to a CLIP text encode, and anima_mod_guidance_quality_tags to a Mod Guidance node when you enable the flag. It also makes a fine bridge node - feed it outputs from other prompt tools and let it normalize the order on the way to the encoder.

Install and gotchas

Part of 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".

The classic mistake is leaving use_anima_mod_guidance on while expecting quality tags in the prompt - with it true, the quality block deliberately vanishes from prompt and only appears on the Mod Guidance output. If your quality tags mysteriously disappear, that's the flag. And the lora_trigger_tags field is one-line only by design; it's for trigger words, not a LoRA list. If you want the same assembly with autocomplete and syntax highlighting, use Anima Prompt Studio - it's this node with a front-end bolted on.