DataverseToCode for VS Code
Keep your Dataverse tables, forms, business rules and form scripts as files in Git: design them in VS Code, test how your forms behave, see an exact plan, push to your development environment, and release to test and production. Query your data with SQL (DQL), chart and share what it finds, and write procedures, views and API calls that run in forms, flows and on a schedule. New here? Open the Get started with DataverseToCode walkthrough on the Welcome page, or Dataverse: Open How-to Guide.

▶ Watch it: the 4-minute tour · the full walkthrough · snippets
▶ Feature videos: the overview (4:13) shows the whole product, and every feature below has its own video (▶), with captions.
What's new
- Query your data with SQL (DQL): queries in
.dql files with results beside them, charts, Export…, Copy as Table and Share as a Dataverse View….
- Procedures: call DQL from form scripts (
await Dql.name()), run it on the server as a Custom API, or on a schedule as a cloud flow.
- DQL views as tables, editable where they can be, and calls to REST APIs from DQL with keys kept in a secured column.
- Virtual tables as code: data sources and connection references in the project, with per-environment values and secrets.
- Offline work: queries and procedures on pulled or seed data, and Save Offline Changes… as a seed scenario.
- Anonymised data: realistic, consistent fakes for personal values in pulled data and query results.
- Guided everywhere: New Dataverse File… creates every kind of file, and Settings… sets a file's header lines for you.
The full list is on the Changelog tab.
A quick look
Design tables, forms and rules as files. The Table Designer edits a table's YAML as a form and a grid; the form designer draws forms as the app does.


Try it and test it offline. Try Form runs a form with its scripts and business rules before anything goes to Dataverse; tests run in the Testing view and your pipeline.

See the plan, push what you choose. Plan shows exactly what would change in your development environment; tick what to push.

Work together on one environment. Claim what you're changing, see who has what, and merge when someone changed the same thing in the portal.


Release chosen stories. Ship just the stories that are ready, to test and production, from VS Code or your pipeline.

Features
Each feature has its own chapter in the How-to Guide (Dataverse: Open How-to Guide), with numbered steps and a check after each one.
Get started
- Install, onboarding and guides: the bundled
dvschema CLI, two walkthroughs, and the How-to and Testing guides inside VS Code. ▶ 4:38
- Connect to an environment: sign in with Microsoft, pick an environment and mark your development one. ▶ 2:36
- New Project: connect and create a project in one step: publisher prefix, solution, CI pipelines, claims and a team snippet library. ▶ 3:43
- Browse an environment: tables by publisher or solution, columns, forms and business rules, and View Data, all read-only. ▶ 3:39
- Pull schema into files: tables, global choices, forms, business rules and form scripts become readable files; only your own things are managed. ▶ 3:11
- New Dataverse File…: one guided place to create a table, virtual table, query, procedure, schedule, DQL view, API, choice, seed scenario, form script, business rule or form. ▶ 3:52
- YAML authoring and validation: completion, hover, go to definition, formula IntelliSense and checks across files as you type. ▶ 2:55
- Table Designer: a table's YAML as a form and a grid of columns, with every column setting, choices, keys and status reasons. ▶ 4:25
- Virtual tables as code: New Virtual Table…, data sources, connection references, per-environment values and secrets, and queries that do what the provider can't. ▶ 4:55
- Form scripts: the JavaScript your forms load, typed for each table, so
getAttribute completes its columns. ▶ 3:26
- Snippets: ready-made form-script code that asks for your table's names; your own and a team library in Git. ▶ 6:01
- Forms: a preview and a designer that draw forms as the app does, new forms, and your app's colours at desktop, tablet and phone sizes. ▶ 4:25
- Design first: draft a form, even for a table that doesn't exist yet, export or share it, then Make it real. ▶ 3:22
- Business rules: flowcharts and plain English, an editor laid out like the portal's, and push that never deletes a rule. ▶ 5:05
Try and test
- Try Form: run a form offline with its real scripts and business rules, and watch what each one does. ▶ 3:50
- Tests:
.test.yaml tests of forms, rules and procedures, generated for you, run offline, in Chromium or online, in the Testing view and CI. ▶ 4:45
Change, review and release
- Plan and push: see exactly what would change, tick what to push, and apply it to a development environment only. ▶ 2:51
- Changes made by others, and merge: see what others changed in Dataverse, with who and when, pull one item, or merge. ▶ 3:16
- Compare environments and drift: a side-by-side diff of a table in two environments, and a check that an environment still matches Git. ▶ 2:06
- Claims and unpublished changes: claim what you're changing on a shared environment, see who has what, and spot drafts in the maker portal. ▶ 4:57
- Solution and Solution Map: what your solution holds, what it depends on, and what's missing. ▶ 2:28
- Releases: managed solutions to test and production, the whole solution or chosen stories, from VS Code or a pipeline. ▶ 4:58
- Git and CI pipelines: pull request checks, a drift check, tag-driven releases and online tests, for GitHub Actions and Azure DevOps. ▶ 1:56
Document
- Relationship diagram: a Mermaid ER diagram of the project that updates as you edit. ▶ 1:05
- Docs and data dictionary: generated Markdown docs that CI keeps current, and an Excel data dictionary of any environment. ▶ 2:03
Test data
- Seed test data: repeatable test records from a scenario, created in dev and removed exactly. ▶ 3:12
- Pull Data: real sample rows, and the records they point at, saved to a git-ignored file for Try Form, offline queries and tests. ▶ 2:13
- Anonymise: realistic, consistent fakes for personal values in pulled data and query results. ▶ 3:14
Query and program your data with DQL
- DQL queries: SQL for Dataverse in
.dql files, with completion, results beside the query, counted writes and saved queries. ▶ 4:21
- Settings…: set a query's, procedure's or view's header lines without remembering them. ▶ 2:45
- Sharing results: Export… to Excel or CSV, Copy as Table for Teams, Outlook and Word, and Share as a Dataverse View…. ▶ 2:50
- Charts: column, bar, line, area, pie and doughnut charts of any result, kept in the query file, saved as PNG or copied. ▶ 2:43
- Working offline: queries and procedures on pulled data or a seed scenario, and Save Offline Changes… as a seed scenario. ▶ 2:48
- DQL procedures:
CREATE PROCEDURE files compiled to procedures.js, so a form calls await Dql.name({ … }); Call Procedure… writes the call. ▶ 4:54
- Server procedures:
-- runs: server makes a procedure a Custom API: all or nothing, callable from flows, canvas apps and the Web API. ▶ 3:49
- Scheduled procedures: Schedule… runs a server procedure by itself, as a Power Automate cloud flow; Choose Connection… picks its connection. ▶ 4:09
- DQL views as tables: a
SELECT becomes a virtual Dataverse table for grids, forms, lookups and flows, editable where it can be. ▶ 5:03
- Calling APIs from DQL:
HTTP_GET and the rest, with SQL Server's JSON functions, from server procedures and views; keys in a secured column. ▶ 3:53
AI assistants
- MCP and AI assistants: Copilot, Claude Code, Codex and Cursor drive the same commands, within the same guardrails. ▶ 2:21
Getting started
Nothing else to install: the dvschema CLI comes with the extension.
- Open the Get started with DataverseToCode walkthrough: it's on the Welcome page (Help → Welcome), or run Welcome: Open Walkthrough… from the Command Palette. Each step has a button.
- Click the Dataverse icon in the activity bar, then Connect to Dataverse. Sign in, pick your development environment and mark it as one; the dialog offers to create a project for it in the same step.
- Right-click your publisher, a solution or a table and choose Pull into Project. Your tables, forms, business rules and scripts become files you can change, review and commit.
- Change things in the designers or the YAML, then Plan to see exactly what would change and Push to Dev to apply it.
- Test your forms: the Test your forms and business rules walkthrough, or Generate Tests in the Project view and then the Testing view. No test-writing experience needed.
- Query your data: New Query… in the Project view's … menu writes a query of a table and runs it beside the file. New Dataverse File… makes procedures, DQL views and APIs the same way.
For more, Dataverse: Open How-to Guide and Dataverse: Open Testing Guide open the step-by-step guides, with screenshots, inside VS Code.
Settings
| Setting |
Default |
Description |
dataverseToCode.cliPath |
empty |
Path to dvschema. When set, that one is used. When empty, a dvschema you installed (on PATH or in ~/.dotnet/tools) is used only if it is at least as new as the bundled one; otherwise the bundled one is used. |
dataverseToCode.openTablesInDesigner |
false |
Open *.table.yaml files in the Table Designer instead of the text editor. |
dataverseToCode.signIn |
vscode |
vscode: sign in with VS Code's Microsoft account. cli: let dvschema sign in with its own browser window and cache. |
dataverseToCode.changes.autoCheck |
true |
Check development environments for changes made by others every few minutes while VS Code is open. |
dataverseToCode.changes.intervalMinutes |
5 |
Minutes between automatic change checks. |
Requirements
Nothing else to install: the dvschema CLI is bundled with the extension, and the .NET 8 runtime it needs is fetched once by Microsoft's .NET Install Tool (installed alongside). A dvschema set in dataverseToCode.cliPath is always used. One you install yourself (dotnet tool install -g DataverseToCode.Cli) is used when it is at least as new as the bundled one; an older one is passed over, with a message saying how to update it.
| |