🐳Clothing Outfit Suggestion

Prompting clothing well is genuinely hard. "Nice dress" gets you a generic dress; the model's training data is full of specific garments - wrap dresses, peplum tops, high-low hemlines - and if you don't name them, you don't get them. 🐳Clothing Outfit Suggestion is a curated answer: pick one garment and a style, and it composes a full outfit as a ready-to-paste prompt string, in English, Chinese, or both. It's not AI - it's a smart dictionary with taste baked in, and for fashion-focused generation that's exactly what you want.

Under the hood it's a hand-built database. The node has 176 garments across eight categories (dresses, tops, bottoms, swimwear, sportswear, underwear, outerwear, special), each stored with its Chinese name and English tag. You tell it the main piece and a style, and it picks complementary items from the right categories, optionally throwing in accessories, and returns both a tag string (for your positive prompt) and a short description. A seed input (‑1 = random) makes it reproducible, so you can iterate on a look instead of rolling dice.

The inputs

• main_clothing - the dropdown with all 176 garments. This is the anchor of the outfit; everything else coordinates around it.
• style - 16 presets from 休闲 (casual), 正式 (formal), 运动 (sporty), 派对 (party), 日常 (everyday), 街头 (street), 波西米亚 (boho), 复古 (vintage), 极简 (minimalist), 浪漫 (romantic), 朋克 (punk), 优雅 (elegant)… Pick the vibe and the suggestions follow it.
• tag_count - 1 to 15, default 5. How many clothing tags the outfit includes. More tags = richer prompt but more chance the model fights itself.
• include_accessories - 是/否 (yes/no). Whether to add accessory tags.
• seed - reproducibility, ‑1 for random.

Outputs and wiring

Two STRING outputs: outfit_tags (the comma-joined prompt-ready tag list) and outfit_description (a human-readable sentence about the outfit). Wire outfit_tags straight into a CLIP Text Encode's positive prompt and you're done. Pair it with the other clothing nodes in this pack - ClothingSelector for manual picks, ClothingSelectorBatch for random variation - and you can cover everything from "I know exactly what I want" to "surprise me."

Install and the fair take

From the comfy_Pond_Nodes pack (Manager search "comfy_Pond_Nodes", or git clone https://github.com/Pondowner857/comfy_Pond_Nodes then pip install -r requirements.txt). It's pure Python data - no models, no downloads, instant.

Fair warning: the garment vocabulary skews heavily toward women's fashion - the 25 dresses alone outnumber every other category, and the options are mostly feminine silhouettes. If your workflow is menswear or gender-neutral, you'll be working around the edges of the database rather than with it. For character/fashion generations in the intended lane, though, the English tags are solid and prompt-tested, which is more than you get from most hand-rolled tag lists.