dvforge

YAML → pac-ready Dataverse table solutions.

Compile compact, human-friendly YAML into a complete Microsoft Dataverse solution — ready to pack with pac solution pack.

[!WARNING] This is pre-production. Breaking changes between minor versions to be expected.

Commands

dvforge build

Compile your input YAML into a Dataverse solution tree.

npx dvforge build --input ./my-solution --output ./output

| Flag | Description | |------|-------------| | --input | Path to your input YAML directory (required) | | --output | Path to write the compiled solution files (required) | | --version | Override the solution version, e.g. 1.2.0.0 | | --unmanaged | Generate an unmanaged solution (default: managed) | | --dry-run | Print output paths without writing any files |

To get it into Dataverse

Pack with the PAC CLI:

pac solution pack -z MySolution.zip -f ./output -p Managed

And import

pac solution import -p MySolution.zip -env $(DV_ENV_URL) -up -pc
# where DV_ENV_URL = https://your-dynamics-url.crm.dynamics.com

Schemas

my-solution/
├── solution.yml
├── optionsets.yml      (optional)
└── entities/
    ├── deal.yml
    ├── contact.yml
    └── transactioncurrency.yml

solution.yml

solution:
  name: MySolution
  display_name: My Solution
  version: 1.0.0.0
  publisher:
    name: TomLumborg
    display_name: TomLumborg
    prefix: tl
    option_value_prefix: 12687

optionsets.yml

optionsets:
  - name: deal_stage          # no prefix needed — dvforge adds tl_
    display_name: deal_stage
    options:
      - label: Talks Open
        value: 1
      - label: Verbal Agreement
        value: 2

entities/*.yml

One entity per file

entities:
  - name: deal                # no prefix needed — dvforge adds tl_
    display_name: Deal
    display_name_plural: Deals
    description:
    ownership: user           # user | organization
    columns:
      - name: name
        type: string
        display_name: Name
        required: true
        max_length: 850

      - name: account
        type: lookup
        display_name: Account
        related_table: account  # no prefix needed

      - name: stage_choice
        type: choice
        display_name: Stage
        option_set: deal_stage  # no prefix needed

    relationships:
      - related_table: account  # the "one" side entity
        lookup_column: account  # the FK column on this entity

Existing (system) Entity

To reference system or built-in entities, add them to the entity list and mark them as an existing table. Use the logical name and display name exactly as they appear in Dataverse. You can also reference custom tables from other solutions by including their prefix.

Once declared, these tables can be used as lookup targets on other tables.

entities:
  - name: TransactionCurrency
    existing_table: true
    display_name: Currency

Column types: string, lookup, choice, datetime, dateonly, int

• choice requires option_set
• lookup requires related_table and a relationship

Prefixes: never write the publisher prefix in your input files - dvforge stamps {prefix}_ on all names at compile time.

Relationships: defined on the entity that holds the lookup column (the "many" side). Only one-to-many is supported.

YAML Schemas: run npx dvforge schema from your project root to generate and configure schemas for your dvforge input files. This enables inline validation, field suggestions, and autocompletion in VS Code.

What gets generated

For each entity:

• entity.yml
• One attribute file per column, plus ~14 system attributes (createdby, createdon, ownerid, statecode, statuscode, etc.)
• Main, card, and quick forms
• 7 standard views (Active, Inactive, My Records, Advanced Find, Associated, Lookup, Quick Find)
• Empty ribbon scaffold
• 6 system relationships + any custom lookup relationships

Globally: publisher.yml, option set files, solution.yml, solutioncomponents.yml, rootcomponents.yml, missingdependencies.yml.

The output directory is wiped and recreated on every build.