Would a compliance reviewer flag this?
Paste a listing, caption, email or ad the way you'd actually post it. It gets the same check every draft in our system gets: the phrases your regulator prohibits, the overclaims, and the figures nobody sourced.
A screening tool, not legal review. Your compliance officer has the last word.
Free · 25 checks a day · no sign-in. We log which phrases matched — never your copy.
Why it doesn’t cry wolf
A phrase list flags everything. This doesn’t, and the difference is the part that took the longest to build.
29 negation words. “We can’t guarantee a closing date” is not blocked as a guarantee — the screen notes it and clears the piece.
Curly apostrophes count. The negator list holds “don’t” with a straight quote. Written the way a word processor actually produces it, it used to tokenise as two words and match no negator — so we don’t guarantee read as a guarantee. There were 222 of them in the corpus when we found it.
A negation has a reach. “With no side effects and permanent results” promises permanence. The word and ends what the no governs, so the second half is still caught.
Warning about a thing isn’t doing it. “If anyone offers to cover your deductible, walk away” is advice against the practice, not an offer of it.
Every one of these exists because the scan got it wrong on real copy first. When the rules are ambiguous it fails toward flagging, never toward clearing — a false alarm costs you a minute, and a miss costs you the thing you bought this for.
And what you have to include
Every checker sells the list of things you cannot say. This is the other half — the identifiers a published piece is supposed to carry. It changes with the industry you picked above.
A missing identifier is not a phrase you can search for, which is why most checkers ignore this half entirely.
What it actually holds
258 prohibited phrases across 7 separate rule sets, plus 29 negation words — so “we can't guarantee a closing date” is not blocked as a guarantee — it is noted and cleared. The gate itself is covered by an automated test suite that runs on every build — it is checked the way the posts are.
| Rule set | Covers | Phrases | |
|---|---|---|---|
| Fair Housing | Real estate | 145 | |
| Reg D | Private offerings | 31 | |
| Reg Z | Mortgage | 26 | |
| Wellness | Medical & clinics | 18 | |
| TCPA | Text messages | 17 | |
| Contracting | Contractors & builders | 16 | |
| General | Everything else | 5 |
Go straight to your industry
Each page is pre-set to that rule set, with real flagged examples and the rewrite for each.
What a flag actually looks like
A trigger term doesn't just get blocked. You get the phrase, why it changed, and clean copy you can publish. This is real output from the checker above — run it yourself with the “Real estate” example.
Flagged
“Perfect family home in a safe Christian neighborhood. No kids allowed in the pool area. Guaranteed to appreciate 20% a year.”
christian neighborhood no kids guaranteed — prohibited for this vertical.20% — figure is not in the approved knowledge base. Source it or cut it.
Publishable
“Three-bedroom home near the Monon Trail, with a community pool and a fenced yard. Recent comparable sales for the street are available on request.”
How the gate works
Checked before you see it
Every piece the system writes is screened against your industry's rulebook before it reaches your queue — not after you publish it.
Flagged phrases are rewritten
A trigger term doesn't just get blocked. You get clean copy, plus what the phrase was and why it changed.
A dated record you hold
Every check is logged, and a one-line receipt rides out in your Friday summary. Once it's in your inbox, the record can't be quietly rewritten later.
Questions
Is this legal advice?
No. It is a screening tool that catches the phrases and claims a reviewer most often flags. Your compliance officer or counsel has the last word, always.
Do you store what I paste?
Nothing is stored except what gets caught. The check runs against the same engine that screens drafts inside our system.
Why do I have to pick an industry?
Because the rules genuinely differ. “Guaranteed” is a problem in mortgage advertising for different reasons than in a medical claim, and Fair Housing language has no equivalent in contracting. Picking the wrong vertical gives you the wrong rulebook.
What is Fair Housing language?
Wording that describes who should live somewhere rather than the property itself — references to family status, religion, national origin, disability or race, including indirect ones like “safe neighborhood” or “perfect for families.” It is the largest rule set here at 145 phrases.
Does it catch numbers?
Yes. Any figure that is not in the approved knowledge base is flagged as unsourced — an appreciation rate, a savings claim, a success percentage. Source it or cut it.
Add-on · $49/month
The gate on its own
This page is the free front of the same engine. The subscription opens all seven rule sets — 258 prohibited phrases, plus the ones you add — and checks a screenshot or a link, not just pasted text.
Already included in Starter, Pro and Teams. This is for people who want the checking without the marketing around it.
This runs on every draft, automatically
The checker on this page is the public front of the gate that screens everything our marketing system writes for you — before it reaches your queue, in your industry's rulebook, with a dated record you keep.
Running a brokerage? The Founding Broker Program screens every listing your agents write — unlimited checks, up to 25 seats, and the dated record is yours. Ten places, $299/month, locked for life.