How the guide is organised
The guide treats the platform as one system centred on Dataverse: apps, automation and deployment artifacts are all metadata records stored there.
The chapter order follows that structure. Chapters 2–3 cover the data model and how you package and move it between environments. Chapters 4–5 cover the apps built on it. Chapters 6–7 cover automation, synchronous before asynchronous. Chapters 8–9 cover data movement and deployment. Chapter 10 documents the environment the guide's claims were verified against.
Chapters are numbered, sectioned and cross-referenced so each one reads on its own. Land here from a search result and you can read a single section and leave with what you came for.
How claims are marked
A guide to a hosted platform cannot execute every claim it makes. So every factual claim here tells you how it is warranted, in three kinds, marked on the page and checked by machine.
-
verified
Run against a real Dataverse environment or the
pacCLI while the chapter was written. You can read the artifact in chapter 10. -
cited
Pinned to Microsoft's documentation, with the specific behaviour or figure named.
-
experiential
Practice judgment from field use, marked as judgment. What it rests on is stated beside it.
An unmarked factual claim is a defect here, and the guide's test suite sweeps the prose for them.
The authoring kit
The guide has its own authoring system — a written spec, a structural checker and a starter skeleton — that holds the chapters to one voice and one design as they land. The spec is the canonical reference; the pages follow it. house-style.md · CLAUDE.md · check.py · skeleton.html