ChatGPT Colour-Blind Outfit Prompts: Use Labels, Patterns and Contrast
“Wear the green one” is only useful if that instruction helps you identify the garment. A wardrobe plan can use names, patterns, light and dark contrasts, shelf locations and item numbers as well as colour.
Ask ChatGPT to plan outfits using the identification cues you prefer, keeping colour names as supplied information rather than an image-based certainty. This can be useful for a colour-blind wearer, or for anyone who finds retailer colour names and screen photographs inconsistent.
Define useful cues before asking for combinations
You do not need to disclose a diagnosis or ask the assistant to simulate your vision. Tell it what makes your own clothes easy to identify: a stripe, a label, a numbered wardrobe list or a lighter-versus-darker description you have confirmed. An assistant should not diagnose colour vision, promise that a palette is accessible to everyone or decide what you can distinguish.
Use written descriptions for the decision. If you request a concept image, keep the selectable outfit list beside it. A generated palette cannot replace the real garment labels or establish exactly how you will perceive an outfit. Retailer names such as ocean green are identifiers, not standardised measurements of colour.
Six prompts for outfits beyond colour names
1. Make an identifiable wardrobe list
ChatGPT, organise these garments using the cues I supply [item names, label colours, patterns, shelf locations and my own light/dark observations]. Give each a stable item number. Do not infer colour blindness, diagnose vision or change my descriptions from a photograph. Ask me which identification cues are useful before proposing outfits, and keep any uncertain colour entries visibly marked unknown.
A stable item number lets you keep the same garment in later prompts even when a retailer uses an unfamiliar shade name.
2. Build combinations with more than one cue
Create three outfits for [actual occasion] using only my numbered wardrobe list [data]. Identify each garment by number, item name and a pattern or location cue, not colour alone. Use my own preferences for contrast. Explain the combinations without universal skin-tone rules, and do not invent garments, measurements or dress-code requirements. Include one familiar option and one small optional change.
A list such as shirt 03, fine stripe, second shelf is easier to act on than an unexplained colour instruction.
3. Keep seller labels separate
Compare these copied product descriptions with my wardrobe [data]. Preserve each retailer’s colour label exactly, alongside pattern and other documented details. Do not claim a photo proves a precise shade or that similar colour names match. Write a short seller question if colour identification matters to the purchase. Keep fabric, size, price and set contents in separate factual fields.
This stops colour uncertainty from spilling into unrelated buying claims.
4. Plan a labelled outfit card
Write an outfit card for my chosen combination [item numbers]. List the top, bottoms, footwear and optional layer with at least two identification cues each. Keep the important information in plain text. Do not use coloured dots as the only labels. Include the event and one practical check I supply, such as whether my existing shoes suit the actual venue.
A card can live in a phone note or on paper. It should remain useful when an image is unavailable.
5. Create an optional concept image brief
Create an editorial outfit concept from this approved written combination [data]. If image generation is available, preserve my supplied garment descriptions and clearly separate imagined styling from actual inventory. Do not label the image a colour-vision simulation or physical-fit test. Keep identifying names in the accompanying text, and leave space for a short title without adding unsupported accessibility claims.
The visual illustrates the combination. It should not become the sole source of the outfit instructions.
6. Review with my actual clothes
Make a final comparison checklist for the garments I have in front of me. Use my preferred identification cues and actual occasion details [data]. Ask me to confirm each item before calling the outfit ready. Do not infer how I see colours, impose a flattering palette or claim a perfect match. Leave the decision to me and record any cue that needs correcting.
Your own feedback improves the next wardrobe list; it does not need to become a medical explanation.
A worked example without a colour-only instruction
Consider a hypothetical wardrobe with shirt 01, solid and on the top shelf; shirt 02, fine stripe and on the middle shelf; and trousers 03, a plain darker pair confirmed by the owner. “Shirt 02 with trousers 03” is an actionable instruction. Adding the wearer’s own contrast preference makes it more personal. The assistant should not invent which shades those garments have or claim that everyone with a named form of colour blindness will prefer the same combination.
A real shirt reference for a factual list
The TryBuy black solid cotton men’s shirt is listed with a Mandarin collar, full sleeves and regular fit. Its product name and solid pattern give identifiers beyond a coloured swatch. Compare the specific chart with a familiar garment; the word black does not establish size or comfort.
Frequently asked questions
Can ChatGPT diagnose colour blindness?
This workflow does not do that. It uses your chosen wardrobe-identification cues.
Should every outfit use strong contrast?
No. Supply the combinations and contrast preferences that work for you.
Do I need an image for these prompts?
No. A numbered text wardrobe can be enough.
Your next useful step
Start with one real garment and one unresolved question. Read the related wardrobe or photography guide for the wider workflow, then check the actual product information before buying or presenting an image as a product reference.
Cover: AI-generated editorial concept, not actual TryBuy inventory, a real photoshoot or physical-fit evidence. Prompt examples are original editorial suggestions. Official sources and any product references were checked on 11 October 2026.