Schema PR Packs — the pack format, and how to verify one

Why publish what we sell. Most consulting firms keep "deliverables" private. We make them public because the differentiator isn't knowing what good schema looks like — it's the engineering follow-through to ship it as a merged PR every week against a moving codebase, measure the citation lift, and adjust. The pack shape is commodity; the operator pattern isn't.

Pack format. Each pack: id, title, priority (high / medium / low), applies-to context, plain-English summary, engine-specific impact note, "before" state baseline, paste-ready "after" block with placeholders, verification step.

How to verify a pack works on your site: (1) drop the after-block into your <head>, (2) replace the placeholder values, (3) run the changed page through validator.schema.org, (4) confirm the types the pack targets now parse with no errors.

Where the blocks come from. Nothing on this site generates them. Each after-block is written by hand against the format above, for the specific page it is going into. Everything Garnet publishes as a downloadable dataset is indexed on /links.

What a schema audit measures →See the GEO subscription →