OriginPrecheck does a focused job on a deceptively simple phrase. You paste a country-of-origin claim before it ships, and it returns a scored pre-check: is this an unqualified "Made in USA", does a qualified claim hold up, are there "assembled" or "designed" traps, and does the copy contradict itself on US versus foreign content. I built it because "Made in USA" is one of the few claims with a fairly explicit FTC standard — all or virtually all domestic — and almost everyone overshoots it.

The standard is stricter than it sounds

The FTC expects an unqualified "Made in USA" to mean all or virtually all of the product is made in the US. The moment you add "assembled in", "designed in", or "parts from", you have changed the claim, and the new claim has its own rules about how much can be foreign. The detection walks the pasted text for these markers and grades the claim type: unqualified, qualified-with-US-emphasis, or foreign-adjacent. It also watches for the internal contradiction — "Made in USA" in the headline and "imported components" three lines down.

This is pattern and consistency work, not opinion. A model will read "proudly assembled in America" and call it a domestic claim. The marker logic separates "assembled" from "made", which is exactly the distinction the rule turns on. A model only writes up the findings.

What the report checks

Four axes. Unqualified-claim scan: did you say "Made in USA" with no qualifier, and does the rest of the copy support that? Qualified-claim handling: if you qualified it, is the qualifier clear and not misleading? Assembled / Designed traps: are those softer words doing the work of a "Made" claim they cannot support? US-versus-foreign contradiction: does the copy fight itself on origin?

A clean report means the wording is internally consistent and the claim type matches the qualifier. It does not mean the product meets the all-or-virtually-all threshold — that is a facts-about-your-supply-chain question.

A concrete example: "Designed in California, assembled in the USA from imported parts" is internally consistent — it qualifies the claim and admits the foreign content. The tool passes the contradiction check because the copy does not fight itself. Swap "assembled" for "made" and the same sentence fails, because "Made in USA" cannot sit beside "imported parts". That single-word swap is the most common error I see, and it is exactly the kind of thing a human skim misses but a marker scan catches. The OriginPrecheck page states that, because the tool catches the wording mistakes, not the sourcing truth.

The limits

It reads text. It cannot inventory your bill of materials or confirm what percentage is domestic. It does not model sector-specific origin rules (some textiles and autos have their own marking laws) or non-US regimes like the EU's. And it is text-only — a "Made in USA" badge in an image is invisible to it.

So the output is decision-support. The report is what you bring to a reviewer: "here is where the claim and the qualifier disagree". If a label ships at volume, the flagged items deserve a real substantiation pass against your actual sourcing. The OriginPrecheck scan just stops the self-contradicting draft from printing.

FAQ

What does OriginPrecheck look for?
Unqualified "Made in USA" claims, qualified claims and their support, "assembled"/"designed" traps, and contradictions between US and foreign-origin statements.

Does it confirm my product is actually all-US-made?
No. It reads wording. Whether you meet the all-or-virtually-all standard depends on your real sourcing.

Is this a legal sign-off?
No. It is a heuristic pre-check for decision-support, not legal advice.

Does it cover non-US origin rules?
It focuses on the FTC standard. It does not model other countries' marking regimes in detail.