🐳Clothing Random Batch

If you're running a batch of generations and want each one wearing something different, typing ten clothing prompts by hand is soul-destroying. 🐳Clothing Random Batch generates up to ten random outfit tag sets in a single run - each one a ready-to-paste prompt - so you can wire them to a batch and get a wardrobe of results in one queue. It's the automation sibling of the Pond clothing family: ClothingSelector is your manual picker, ClothingSelectorSimple is number-driven, and this one is the "surprise me, ten times" button.

The engine is a weighted random sampler over the same curated 176-garment catalog the other clothing nodes share. It draws garments per category, weights the categories (you control the odds), optionally constrains the draw to a style preset, and guarantees certain tags appear. A seed input (‑1 = random) makes the whole batch reproducible, which matters more than it looks - being able to freeze a set of outfits you liked is how you iterate without losing work.

The inputs

• batch_count - 1 to 10, default 3. How many outfits you get. The outputs mirror this: batch_1 through batch_10 (unused ones come out empty).
• tags_per_batch - 1 to 10, default 3. Garments per outfit. 3 is a sensible outfit core (top/bottom/dress + something); push higher for richer prompts.
• category_weights - a STRING like 连衣裙:0.2, 上衣:0.2, 下装:0.2, 泳装:0.1, .... This is the balance knob: raise 下装 and you get more pants-centric outfits, zero a category and it never appears.
• style_preset - 随机 (random), 休闲 (casual), 正装 (formal), 运动 (sporty), 性感 (sexy), 日常 (everyday). Narrows the draw to a vibe.
• ensure_tags - a comma string of tags you always want, e.g. sneakers, denim jacket. The node guarantees these land in every outfit. This is the killer input for consistency.
• seed - reproducibility.

Outputs and wiring

Ten STRING outputs, batch_1 … batch_10. The clean pattern: feed each into a batch prompt system (or a prompt-schedule node) driving a batch KSampler, so image N gets outfit N. Because each output is just a comma-joined tag string in the same format as the other clothing nodes, they're drop-in compatible with the rest of your prompt plumbing.

Install and the fair warning

From comfy_Pond_Nodes (Manager search, or git clone https://github.com/Pondowner857/comfy_Pond_Nodes + pip install -r requirements.txt). Pure Python, no models.

The warning is the family one: this catalog is women's-fashion-heavy, so batches skew that way. And "random outfit" means the same garment vocabulary recombined - you'll get variety in combinations, not surprise categories. But for generating a quick spread of fashion prompts to feed a batch, this is genuinely the most useful of the four clothing nodes in the pack.