Every restriction we check links the regulation it came from.
Tamats screens ingredients and claims against a compiled corpus of government regulations and retailer clean-beauty standards, and every finding links the primary source it was compiled from. Findings are computed against the rule set rather than generated as prose, so an ingredient with no matching rule returns 'review' — an honest gap — instead of a fabricated pass. The corpus is a screen, not a legal opinion: it tells you what to look at and shows you the text, and a regulatory professional still makes the call on a borderline case.
Most compliance work is not hard, it is wide. The same ingredient is fine in one market, capped in another, and quietly excluded by a retailer's own clean standard that has nothing to do with any government. Getting that wrong is not a small mistake — it is a reformulation after tooling, or a delisting after launch. The screen exists to make the wide part cheap so the hard part gets the attention.
What you do with it
Per-ingredient status with its source
Each ingredient resolves to banned, restricted, flagged or review for a chosen market or retailer standard, with the concentration limit where one exists and a link to the primary text it was compiled from.
Side-by-side market comparison
Screen one formula against several markets and retailer standards at once and read the differences as a matrix, which is the shape the decision actually has when a product is launching in more than one place.
Marketing claims scanner
Ad and label copy is scanned against deterministic, cited claim rules — the phrases that turn a cosmetic into a drug, or a clean standard into a delisting — with the rule text shown, not summarised.
Your own standards, alongside the public ones
Upload a retailer's or a manufacturer's own restricted-substance list and it is screened next to the compiled corpus, so an internal standard is not the one thing checked by hand.
What it computes over
—A compiled rule corpus built from primary regulator and retailer publications, regenerated by script rather than hand-edited
—The CosIng-derived ingredient dictionary, with CAS numbers where the source provides them
—Deterministic marketing-claim rules, each carrying the text it enforces
—Compliance documents your organisation uploads, searched alongside the public corpus
Where it stops
A screen is not a regulatory opinion. It narrows what needs a professional's attention; it does not replace the professional.
An ingredient with no matching rule returns 'review'. That is a gap in the corpus being reported honestly, not a clearance.
Coverage is what has been compiled from primary sources. Markets outside the compiled set are not silently approximated from adjacent ones.
No safety assessment. Screening an ingredient against a limit is a different thing from assessing a product's safety, and only the first happens here.
Who picks it up from here
A regulatory affairs professional or a qualified safety assessor makes the call on anything the screen returns as restricted or review — with the primary text already pulled up rather than still to be found.
Questions
What happens when an ingredient is not in the corpus?
It returns 'review'. The product never converts an absence of data into a pass — an unmatched ingredient is reported as unmatched, with the target it was screened against named, so you know exactly what was and was not checked.
Are retailer clean standards really regulations?
No, and they are labelled separately. A retailer's clean standard is a commercial condition of being listed, not law — but it delists products just as effectively, so it is compiled and screened alongside the government rules rather than left to a PDF someone forwards.