Questions, answered plainly
The short version: we generate only from your verified facts, every line shows its source, and a line we can't trace blocks export.
A general chat tool will write “5+ years” or “led a team of 4” if it sounds right, because nothing checks what it writes against what you've actually done. That isn't a bug in those tools; it's not their job.
Here, it's the whole job. Every line cites the facts it was written from. Code checks each line against those facts — numbers, tools, employers, anything you said you don't have — and an AI reviewer reads each one too. A line we can't trace is flagged, and it blocks export until you back it up or take it out. See the fictional example.
You don't have to take its word for it. Every line it writes cites the facts it came from, and you can open any line to see them.
Each line is checked against those facts: by code, which is strict about numbers, tools, employers and anything you said you don't have, and by an AI reviewer reading for meaning. A line we can't trace is flagged, and it blocks export until you back it up with a fact or take it out. The checks can miss things — read what you send.
You provide the truth; we do the bookkeeping. You keep a list of facts about your work — typed in, or imported from LinkedIn or an old résumé and confirmed line by line. For each posting, you say which of its requirements you actually have.
Then Crane writes the wording: a résumé and cover letter tailored to that posting, using only those facts. You see where every line came from, and you can edit anything. What you export is yours to read and send — we never submit an application for you.
We ask you. For each thing the posting asks for, you say whether you have it. If you do, you back it with a fact you already have, or describe it in your own words and it becomes one.
If you say no, or don't answer, it's left out: never mentioned, and never hinted at with words like “exposure to”. The gap report lists what was left out and why, so you see the gaps before you apply.
No. There's no job board, no search, and no recommended jobs. Every application starts with a posting you brought. What we do help with is learning from the applications you make: you can record what happened to each one, and see your own counts.
No. It runs out; it isn't worse. Free has everything that makes a document trustworthy and usable: the checks on every line, seeing what each line is based on, the gap report, and export to PDF, Word and plain text. It includes 10 posting analyses and 4 generations a month.
Pro raises those caps and adds analysis on top: 6 things, listed on the pricing page. Fixing anything — restoring an earlier version, merging or retiring a fact — is free on every plan.
It's private to your account: your facts, postings and documents can only be read by you, and the database enforces that, not just the app. There are no ads and no tracking scripts.
To analyze a posting and to write and check your documents, the text involved is sent to Anthropic's API (Claude), which does the writing and reviewing. On Pro, the name of the company you're applying to is sent to Brave Search for company research. When billing launches, payments go through Stripe; we never see your card.
You can delete any fact or application yourself, at any time. Deleting your whole account isn't a button yet; a contact address for that will be listed here before launch.
No. There's no match score, no percentage, and no ATS grade. A score would be a guess about how an employer's software ranks people, and no one outside that employer can see that software.
Instead we show what the posting asks for, which of your facts show each thing, and what's missing — so you can decide whether to apply, and what to say.
We don't show an “ATS score” or a match percentage, because no one outside an employer can know how their software ranks anyone. What we show instead is plain: what the posting asks for, which of those your facts show, and what's missing. Exports are deliberately plain — one column, real text — so they read cleanly everywhere.
They can. The code checks are strict about numbers, names and tools; the AI reviewer reads for meaning, and a reviewer can be wrong. That's why you see what every line is based on, one click away, and why you should read what you send. The checks catch the common inventions; they don't replace your eyes.