Why do AI resume builders invent experience?
Not because they are badly built. Because of what they are.
A language model asked to make your resume match a job description is being given two instructions that quietly conflict. One is "match this job." The other is "stay true to this person." When those two things can both be satisfied, the model rephrases your real work to speak the job's language, which is genuinely useful. When they cannot, something has to give, and the model has no way to know which instruction is the important one.
So it fills the gap. It reaches for the Kubernetes experience the job wants and you do not have. It rounds "helped with" up to "led." It attaches a number to an achievement that never had one. The output is fluent, confident and completely plausible, because fluent and plausible is exactly what the model was trained to produce. It is not lying in any way it could detect. It ran out of facts and kept writing.
An AI does not fabricate because it is dishonest. It fabricates when it is under pressure to produce and has run out of things it actually knows.
That framing points at the fix. If fabrication happens when the model runs out of facts, then you solve it by giving the model more facts and removing its permission to write without them. Not by asking it nicely in a prompt.
What is the Honesty Engine?
The Honesty Engine is the set of checks a tailored resume has to pass before JobShifu will produce a document. A tailored resume is treated as a build artifact, the way compiled software is: it either passes and exists, or fails and does not.
One check blocks. Five report.
| Check | What it catches | Result |
|---|---|---|
| Provenance | A line citing no confirmed source, or citing a source that does not exist. This is the non-fabrication guarantee. | Fails the build |
| Role integrity | Real work filed under the wrong employer. A line sitting under one job whose evidence all belongs to a different one. | Warns |
| Jagged metrics | A number that appears in no cited source, or a timespan that does not fit the role it sits in. | Warns |
| Scope | A line claiming more reach or ownership than the source behind it supports. | Warns |
| Entailment | Substance the citation never contained: a tool, outcome or responsibility absent from every source the line cites. | Warns |
| AI tells | Buzzword and weasel-word density. The texture that makes a resume read as machine-written. | Advisory |
The split is deliberate, and it is worth being precise about which check carries the guarantee.
Provenance is exact and deterministic. It is ordinary code reading the citations attached to each line and confirming they point at confirmed Vault units. It cannot be argued with, it has no opinion, and it returns the same verdict every time. That is what makes it safe to block on.
Scope and entailment use a model, so they advise rather than block. Judging whether a line overstates its source is a judgement call, and a model making judgement calls will sometimes be wrong. Blocking your resume on a check that can be wrong would be its own kind of dishonesty. So they flag, you decide, and the deterministic check is the one holding the floor.
How does tailoring actually work?
The short version: JobShifu proposes, you approve, and the gate runs last.
-
Your Vault holds confirmed facts
Roles, projects, achievements, skills. You can import a resume to start it, but nothing becomes citable until you confirm it. Unconfirmed material is invisible to the tailoring step, which means the model cannot build on something you have not stood behind.
-
The model proposes changes, it does not write your resume
JobShifu reads the job description and returns a set of specific edits: keep this line, rewrite that one, drop this, combine those two, add one built from these Vault entries. Every proposal names the confirmed units it draws from. Proposals citing anything unconfirmed are dropped before they reach you.
-
You approve every change, card by card
Each proposal is a card showing the original, the proposal, and the evidence behind it. Accept, edit, or reject. There is no "accept all" that skips the reading, and that is on purpose: the review is where you catch the line you would not want to defend, and it is also how JobShifu learns what you actually sound like.
-
The gate runs before the document exists
Your approved changes are assembled and the full Honesty Engine runs on the result. Provenance passes or the build fails. Only on a pass does a finished resume version exist to preview, download or attach.
What happens when a resume fails the check?
You get told which line failed and why, and no document is produced.
This is the part people find surprising, so it is worth stating plainly: a failed build is not a degraded resume with a warning on it. It is the absence of a resume. There is nothing to accidentally download, nothing to forget you were warned about, nothing sitting in your documents folder six weeks later that you attach to an application without remembering the banner you dismissed.
A warning you can click past is a warning that eventually gets clicked past, usually at 11pm on the night before a deadline. That is exactly when you least want the safety net to be optional.
Does this mean my resume gets worse?
It means your resume stops borrowing from experience you do not have, which is a different thing.
Most of what makes a tailored resume work is not invention. It is selection and language: choosing which of your real achievements to lead with for this specific job, and describing them in the terms this employer uses. Both of those are fully available inside the constraint. JobShifu will happily rewrite "reduced page load time" as "improved frontend performance" when the job description asks for the second phrasing, because that is your real work in the job's vocabulary.
What it will not do is give you the distributed systems experience the job wants and your Vault does not contain. And here the constraint is doing you a favour, because that line was never going to survive the interview anyway.
There is also a compounding effect that only works if the facts are real. Every job you read teaches JobShifu something about what employers keep asking for, and the gaps it finds turn into concrete things you can go and actually do. A Vault built from real work gets genuinely stronger over time. A resume built from invention has to be maintained as a story you remember.
Passes the Notepad test
Paste any bullet into a blank document and you can still explain where it came from, because it came from something you confirmed.
Survives the interview
Every line traces to real work, so there is no bullet on the page you cannot talk through for two minutes.
Frequently asked questions
Will an AI resume builder lie on my resume?
Most will, given the chance. A general-purpose language model asked to make a resume match a job description has no way to tell the difference between rephrasing something you did and inventing something you did not. When it runs short of facts it produces fluent, plausible, unverifiable text, because that is what it was built to do.
JobShifu removes the opportunity instead of asking the model to behave. Every line of a tailored resume must cite a confirmed unit in your Vault, and a line with no source fails the build before a document is ever produced.
What does "no source, no line" actually mean?
Every line on a JobShifu resume carries a pointer back to the specific Vault entries it was built from. Before the finished document is assembled, the provenance check verifies each line cites at least one confirmed source and that every cited source exists. If any line fails, the build fails.
This is not a warning banner you can click past. There is no document on the other side of a failed check.
Does JobShifu write my resume for me?
No, and the distinction matters. JobShifu proposes changes and you approve them one at a time. Each proposal is a card showing what it wants to change, what it would become, and which of your confirmed experiences it draws from. You accept, edit or reject each one.
A resume you never reviewed is a resume you cannot defend in an interview, so the review step is deliberately not skippable.
Can I still exaggerate if I want to?
You can edit any line, and JobShifu will not silently rewrite what you wrote. What it will do is tell you when a line drifts past what your own evidence supports. The scope and entailment checks flag lines claiming more than their cited sources contain, jagged metrics flags numbers that appear in no cited source, and the title check flags a display title reading above your Vault seniority.
These surface as warnings rather than blocks, because the line is yours. The one thing that cannot happen is an invented line arriving without you noticing.
Is the honesty check just another AI reviewing the output?
The check that blocks is not AI at all. Provenance is exact, deterministic code: it reads the citations on each line and verifies they point at confirmed Vault units. It cannot be talked out of a verdict, and it returns the same answer every time.
Two of the six checks do use a model, scope and entailment, and those are advisory precisely because a model can be wrong. The guarantee rests on the deterministic check, not the model-backed ones.
Why does this matter if the resume gets me the interview anyway?
Because the interview is where an invented resume gets discovered. A line you cannot talk about for two minutes is a line that will cost you the offer, and you will not know which line it was until you are sitting in front of someone asking about it.
Every bullet JobShifu produces traces to something you actually did, so there is nothing on the page you cannot walk through out loud.