Dress from what you already own.
Worn is a private wardrobe. Keep every piece in one place, get an outfit for today's weather, log what you wear, and ask ChatGPT or Claude about your own closet.
Connects to ChatGPT and Claude as an MCP connector.
Closet
15 pieces
Tap a category to filter the closet.
Every piece, filed properly.
Add photos straight from your camera roll. Set the category and fabric, look up the brand and product, and mark anything at the cleaner so it stays out of suggestions until it is back.
- Category and fabric edits in a tap
- Brand and product lookup, with logos
- Optional background removal, originals kept
Camel blazer
Outerwear, worn 18 times, last on September 16
Fabric
Today, decided.
Worn checks the forecast, skips anything unavailable, and lays out a look from your own closet. Shuffle until it feels right, save the ones you like, and log the one you wear.
Today
Friday, September 18
54°
Light rain in Brooklyn
- High
- 58°
- Low
- 47°
Suggested from your closet
Merino under wool for a 47° evening.
Look 1 of 4
Camel blazer
Wool
Oat crewneck
Merino
Stone trousers
Cotton twill
Bone sneakers
Leather
Left out: Olive jacket, marked at the tailor.
Talk to your closet.
Add Worn to ChatGPT or Claude as a custom connector. Your assistant can then search, suggest and log from your real wardrobe through fifteen tools, instead of guessing.
- 1Add the connectorPaste Worn's MCP address as a custom OAuth connector.
- 2Sign inUse the same Worn account you use in the app.
- 3Review and connectRead the consent screen, then approve.
Things to ask
- “Find my linen shirts.”
- “Let me upload a new batch of clothes.”
- “Suggest an outfit for today's weather in Brooklyn.”
- “I wore these two pieces today, log them for September 18, 2026.”
An illustrative exchange. The tool names are Worn's own.
Bring in the whole rail.
Upload a batch and review it in one pass. Fix categories and fabrics, choose what comes in, and let Worn catch a photo it has already seen. Imports are retry safe, so nothing lands twice.
Review 6 photos
Set a category where it is missing. Untick anything you want to leave out.
Chalk shirt
Poplin
Sage knit
Cotton
Cocoa trousers
Corduroy
Ivory dress
Linen
Ecru oxford shirt
Same photo is already in your closet.
Oxblood tote
Leather
One piece still needs a category.
What you actually wear.
Log an outfit in a tap, or just tell your assistant. The favourites and the forgotten both become obvious. Undo is guarded: it only reverses an entry if nothing has changed since.
September
15 of 18 days logged
September 17
- Olive linen shirtWorn 9x in total
- Ink wide trousersWorn 15x in total
- Ink toteWorn 20x in total
- Most worn
- Bone sneakers, 7 days
- Untouched this month
- Olive jacket
Connect your wardrobe
Only connect an app you recognize. Claude is asking for access to your Worn account.
- See your clothes, looks and wear history
wardrobe:read - Add pieces, edit availability and log outfits
wardrobe:write
You sign in on Worn, not inside the chat, so the assistant never sees your password.
Private by design.
Every account has its own wardrobe, and ownership is enforced by the database. Assistants only get in when you say so.
- Consent, by name
- You approve each assistant yourself, on a Worn screen that lists exactly what it can do.
- PKCE
- Every sign in carries a one time secret, so an intercepted code is useless on its own.
- Dynamic client registration
- ChatGPT and Claude register themselves with Worn. There are no keys to copy or paste.
- Your photos stay private
- Photos live in private storage, and the database itself checks that every piece belongs to you.
Built with.
A real product, not a concept. Conflict detection, retry safe imports and logging, and guarded undo sit underneath everything you see above.
- Expo and React NativeThe Worn app itself
- A stateless Node APIWardrobe logic and the MCP server
- SupabaseAuth, Postgres, and private photo storage
- VercelHosts the cloud app and the API
- Model Context ProtocolFifteen tools for ChatGPT and Claude
- SerpApi and BrandfetchProduct lookup and brand logos
A note from the maker
Most mornings the question is not what to buy. It is what to wear, from clothes that are already in the closet.
I built Worn on my own in September 2026 to answer that: a private place to keep your clothes, a suggestion that respects the weather, and a way to ask the assistant you already use. No fitting rooms, no avatars, no feed. Just your clothes, kept well.
Muhammad Bin Sohail
Designed and built Worn, end to end