Three colors, one rubric.

Every product gets one of three verdicts: green, yellow, or red. The rating is auto-computed from the product's ingredient list — not from taste tests, not from marketing, not from how much we liked the bottle. If you don't like a rating, you can trace it back through the ingredient rubric to the exact row that triggered it.

The rules

• Green — every listed ingredient is in our "clean" category. Nothing watched, nothing flagged.
• Yellow — one or two watched ingredients (common additives that studies are mixed on: sucralose, soy lecithin, artificial colorings).
• Red — any flagged ingredient (ingredients with clear adverse public-study signal), or three or more watched ingredients stacked on one label.

What we don't measure

We don't run lab tests. We don't measure heavy-metal contamination, actual vs. claimed protein content, or microbial load. Labdoor and ConsumerLab do that and do it well. If you want chemistry, go there. We review what a brand prints on the label.

Why this is the whole game

A label is a promise a brand makes to a regulator and, more importantly, to you. What they put on it is what they're defending in writing. Reviewing the label holds brands to their own word. Lab testing catches lies; label-reading catches the clearly stated compromises.