← Back to Changelog

Store Dimensions

Store dimensions panel listing targetable fields with coverage percentages next to a write validation panel rejecting an unknown field and suggesting the correct one

Building merchandising rules, blocks, and sort orders with an AI assistant used to mean guessing which context fields your store supports. The new store-dimensions MCP tool answers that directly: every dimension and contextual field your store has seen, with coverage percentage, unique value counts, and when it was last observed.

Pass a field name to sample real values seen in production, or filter by group, like Geography or Marketing, to narrow things down. Writes now validate against the same list, so a typo like context.geo.provnce gets rejected up front, with context.geo.province suggested in its place, instead of saving successfully and never matching a shopper.

Learn more