VNCCS Prompt Concat

Four string inputs - a, b, c, d, each an empty single-line string by default - one separator field (default ,), and one output: all four strings joined together with that separator between them. That's the entire node. No conditional logic, no filtering out empty slots, no reordering - just straightforward concatenation in the order you wired it.

Why this exists in a character-creation pack

VNCCS's Character Creator and Clothes Designer split character description into distinct categories on purpose - the README specifically calls out head and face as their own fields, separate from general body or clothing description, because those two feed directly into the face-detailer pass during emotion generation later in the pipeline. Keeping "glasses" in the face field rather than buried in a general description isn't cosmetic; it's what lets a downstream node target the right region.

That kind of categorized input needs recombining into a single flat prompt string at some point before it reaches a text encoder, and that's the gap Prompt Concat fills. Feed it separate strings for, say, character tags, outfit tags, pose description and a style suffix, and it hands back one clean prompt with your chosen separator between each piece - comma being the obvious default for tag-style prompting, though nothing stops you from setting it to a space, a newline, or anything else if you're feeding a natural-language encoder that doesn't want comma-separated tags.

Where it fits

Sits between whatever's building your tag categories (typed manually, or produced by one of VNCCS's Wizard features) and whatever consumes the final string - most likely VNCCS QWEN Encoder's prompt input, or a VNCCS_MultilineText node if you want to review the combined string before it goes further.

How to install it

Comes with the main pack:

• ComfyUI Manager - search VNCCS - Visual Novel Character Creation Suite, install, restart.
• Manual - cd ComfyUI/custom_nodes && git clone https://github.com/AHEKOT/ComfyUI_VNCCS.git, then cd ComfyUI_VNCCS_Utils && pip install -r requirements.txt, restart, and let Comfy Manager grab any missing dependencies.

Common issues

Doubled-up separators, or a separator floating at the start or end of the result. Because the node concatenates unconditionally, an empty slot (say, d left blank because you only had three categories to combine) still gets a separator inserted next to it - you'll end up with something like tag1, tag2, tag3, with a trailing comma and space, rather than the separator being smart enough to skip empty fields. If that trailing punctuation matters for whatever's reading the prompt next, either leave unused slots genuinely at the far end of the chain (d rather than b) or clean it up with a text-processing node afterward.

Prompt reads oddly to a natural-language model. If you're feeding the result into an encoder built for prose rather than comma-separated tags - Qwen-Image-Edit generally prefers plain instructions over tag soup - a comma separator produces exactly the disjointed, tag-like phrasing you don't want. Switch the separator to a space or write connective words directly into your category fields instead.