Grounded in what you own
Ask your collection anything.
Which bottles am I neglecting? What should I wear tonight? Which samples deserve a full bottle? Build me a week from what I already own. Six flows answer six decisions — and every one of them starts by handing the model your collection, your wear history, your ratings and the weather at your door, then comes back with its reasoning attached.
What's your vibe?
Set the mood — or skip to use defaults.
Ask, then check the answer
Describe the moment. Interrogate the pick.
Before it ranks anything, the suggester runs more than twenty read-only queries against your own data: collection, wear history, ratings, skips, occasion and time-of-day patterns, layering combos, live weather, and anonymized signals from wearers whose collections resemble yours.
Three steps, all optional — set Vibe, Scene, and inventory intent. Skip them and it fills the context itself from the time of day, the season, and the live forecast.
Ask in your own words — "cozy but not gourmand" keeps the mood and penalizes the gourmand notes. "Something like my Oud Wood" makes it look your bottle up first. If the model is unreachable, a keyword parser takes the ask instead of failing.
Outfit matching — drop in a photo and picks stream back as they are written. What it reads is fabric weight and color temperature, not the brand on the shirt.
Every pick opens — open one for its match score, its weather read, and the reasons behind both. Log it, skip it with a reason, or set it aside as a sample to try first.
“Something smoky but polished for a cold dinner downtown.”
Cloudar
Better World Fragrance House
The darker woods hold up in cool air, while the polished drydown fits dinner without crowding the room.
Talk to your shelf
The same questions, from ChatGPT or Claude.
Fraghab is in ChatGPT's plugin directory and available as a Claude connector, so the collection you built here answers from inside a chat you were already having — and the assistant can write back to it.
It reads the real thing
Your bottles and decants with remaining millilitres, ratings, wear counts and last-worn dates, plus your shelves and layering combos. Not a popularity list — the shelf you actually have.
It can act, not just answer
Add a bottle, log a wearing, correct one logged wrong, set a decant's verdict, promote a decant to a full bottle, build a shelf or a combo.
Your account only, and you approve it
You approve every permission, the assistant asks before it changes something you already had, and there is no tool that deletes anything — deletion stays in the app.
Disconnect whenever you like
From Integrations in your account. To cut every connected assistant off at once, change your Fraghab password: connector access is tied to your password version.
Plan past today
A week and a trip are different problems.
The weekly planner works off your calendar and a seven-day forecast. Travel mode is a different job: it has to know the destination's weather and what will clear security.
Weekly planner — set each day's occasion, or let a connected calendar fill them in. It then balances the week's weather against your rotation, the variety across seven days, and the bottles you never reach for.
Lock, swap, log — hold a day's pick in place, take an alternate, read the score behind either, then log the wearing when the day arrives.
Travel mode — search a destination and set your dates. Back comes a day-by-day rotation off the forecast, with substitutes, a travel-friendly count, and a flag on anything over the 100ml carry-on limit.
Weekly planner
Plan your week, then get your picks.
Travel mode
Pack the perfect rotation for your trip.
Decide the next bottle
Look at the gap, not the want.
What to Buy is a research agent. It maps the overlap in what you already own, reads your budget and your wishlist, then searches and opens real retailer pages before it recommends anything.
Give it the brief — name what you are after, or hand it your wishlist and a budget and let it work out where the gap is.
Prices it actually loaded — the figures come off the product pages the agent opened, not from memory, and a candidate it cannot link to a real page is dropped before it reaches you.
Refine, and keep the trail — push it cheaper, bolder, or toward less overlap. Save a candidate to your wishlist, revisit a past run, or take Try Before You Buy when a sample is the cheaper answer.
Your collection is strong in sweet amber. The better addition adds dry woods and incense without repeating the same dense vanilla profile.
Gambit
Mind Games · est. $168
Best balance of contrast, winter versatility, and price. The incense profile fills a real gap without crowding what you own.
It keeps score
The suggester grades its own picks.
A recommendation nobody checks is just an opinion. Fraghab records what it suggested each day and compares that against what you actually wore.
An accuracy record you can see
After three days of picks, the suggest page shows how often you wore one of them. It sits on the screen, not in an admin panel.
Skips carry weight
Turning a pick down with a reason pushes that bottle back, and the penalty halves every five days instead of blacklisting it forever.
The same engine on your phone
The iOS and Android apps call these flows through the same API the web app uses, so the plan you built on a laptop is the plan you have on the plane.
Important information
AI suggestions, collection analytics, and retailer price information are informational tools, not purchase, financial, medical, or safety advice. Verify product details, pricing, availability, authenticity, and suitability directly with the relevant seller or brand before making a decision.
Fraghab does not process peer-to-peer payments, provide escrow, or authenticate people or products. Community content reflects its authors' views. If you choose outfit matching, your uploaded photo is processed to produce the requested suggestion as described in our Privacy Policy.
Ask your collection
It is only as good as what you have cataloged.
Catalog a few bottles and the suggester has something real to reason over. From there, every wearing you log teaches it which of its picks you actually reach for.
Free · No card · Core collection tracking
