Custom Module Rendering

Text size

Custom Module Rendering

Add stackgen.yaml to a custom Terraform module to tell StackGen how that module should render in the UI.

The file goes in .stackgen/stackgen.yaml (or .stackgen/stackgen.yml) at the module root. StackGen used to use properties.json and metadata.json for the same purpose. Those are now one YAML file.

Use this when you build a module from an abstract or work in the Module Editor. Autogenerate stackgen.yaml creates or refreshes the file. From there you set the canvas label and icon, how inputs appear in the configure panel, how modules connect on the topology, and how existing cloud resources are discovered for import.

You can validate stackgen.yaml schema keys and structure while you author or update custom modules:

• API / CLI: Validate schema keys programmatically during custom module development. This cut checks schema keys for how the module is presented on the canvas. It is not full semantic or runtime validation of every canvas behavior. See v2026.7.7.
• appStack-owned Module Editor: When you create or edit stackgen.yaml for an appStack-owned module, StackGen validates incorrect keys. You get an error toast, and an error count appears in red next to stackgen.yaml in the side pane so you can fix schema issues before you publish.

Quick Start

1. Create .stackgen/ in your Terraform module root.
2. Add stackgen.yaml.
3. Commit and push the module.
4. Import or refresh the module in StackGen.

What Goes in the File

Block	What it does
representation	Icon, label, and description on the Topology canvas and side panel
container	Container behavior: delete rules and how other modules connect in
variables	Labels, controls, and validation for each Terraform variable
variable_groups	Sections in the configure panel
connections	Predefined wiring between modules (outputs to inputs)
discovery	How StackGen finds an existing cloud resource to import

Set version: "1.0" at the top of the file.

Representation

The representation block sets how the module appears on the Topology canvas and in the configure resource side panel.

You can set a short description, an icon, side panel label and icon, and how the canvas node is titled. Use node.label.static for the default title and node.label.template with ${variable_name} once the user fills in a variable. Set node.display_type: container when the module should render as a container. node.label_attribute names the variable the template reads (for example name or bucket_name).

Example

Add this under representation in .stackgen/stackgen.yaml:

representation:
  description: |
    DynamoDB table module with container, discovery, and connection metadata.
  icon: aws-dynamodb-table
  side_panel:
    label: DynamoDB Table
    icon: aws-dynamodb-table
  node:
    label:
      static: DynamoDB Table
      template: DynamoDB Table - ${name}
    display_type: container
    label_attribute: name

The above configuration will:

• Show the description and aws-dynamodb-table icon on the canvas.
• Title the side panel DynamoDB Table.
• Show DynamoDB Table on the node, then DynamoDB Table - plus the name value after the user sets it.
• Render the node as a container.

After you edit the file, commit, push, and re-import or refresh the module so StackGen picks up the changes.

Click to view representation fields

Field	Description
description	Short summary shown in the UI
icon	StackGen icon id (for example aws-dynamodb-table)
side_panel.label	Title in the configure resource panel
side_panel.icon	Icon in the configure resource panel
node.label.static	Default node title on the canvas
node.label.template	Node title with ${variable_name} placeholders
node.display_type	Canvas node type (for example container)
node.label_attribute	Variable name used with the template

Container

We've added native child module support for custom container modules. Custom containers can mirror the nested visual behavior of built-in VPC and subnet configurations on the Topology canvas.

Previously, nested child modules and parent/child attribute mapping were limited to platform-built containers. Now, you configure the same nested canvas behavior in the custom module's stackgen.yaml.

How to configure

1. Set node.display_type to container under representation.node (see Representation).
2. Add a container block with type, on_delete, and connection_rules so parent and child modules map attributes when a child is placed on a parent.

Key Features

• Child Module Support - Nest child modules on the canvas the same way you group built-in networking containers.
• Connection Rules and Attribute Mapping - Define which parent and child modules can connect (for example a VPC parent and a subnet child). When you drag a child onto the parent, valid attributes (such as a VPC ID) map automatically from those rules.
• Deletion Types - Control what happens when the parent is deleted with on_delete:
  • detach - Nested child modules stay on the canvas.
  • cascade - The parent and all nested children are removed together.

note

Visual grouping and attribute mapping rely on your custom connection rules for canvas continuity. This is not a Terraform restriction. StackGen will allow uncommon connections (for example connecting an S3 bucket to a route table) when your YAML rules dictate it.

For YAML template examples or implementation questions, contact support@stackgen.com or your Customer Service Manager.

See v2026.6.12.

Example

container:
  type: networking
  on_delete: cascade
  connection_rules:
    - source_module_name: aws_vpc
      connection_type: DATA
      mapping:
        source_attribute: vpc_id
        target_attribute: vpc_id
        data_type: string

The above configuration will:

• Treat the module as a networking container.
• Remove nested children when the container is deleted (cascade).
• Let an aws_vpc module connect with a DATA link from vpc_id to vpc_id.

Click to view container fields

Field	Description
type	Container category (for example networking)
on_delete	cascade removes children on delete; detach keeps them
connection_rules	List of allowed inbound connections
connection_rules[].source_module_name	Module that connects into the container
connection_rules[].connection_type	Connection label (for example DATA)
connection_rules[].mapping.source_attribute	Attribute on the source module
connection_rules[].mapping.target_attribute	Attribute on the container module
connection_rules[].mapping.data_type	Expected type (for example string)

Variables and Variable Groups

Each key under variables matches a variable in variables.tf. Use it to set the field label, description, type, ui_control, options, and validation. For object and list types, nest fields under attributes with the same shape.

Use variable_groups to group variables into tabs or sections in the configure panel.

Example

variables:
  billing_mode:
    label: Billing Mode
    description: The billing mode of the DynamoDB table.
    ui_control: dropdown
    options:
      - PROVISIONED
      - PAY_PER_REQUEST
    validation:
      required: true
    type: string

  server_side_encryption:
    label: Server-Side Encryption
    type: object
    default_editor: form
    variable_label: Encryption block
    attributes:
      enabled:
        label: Enabled
        type: bool
      kms_key_arn:
        label: KMS Key ARN
        type: string

variable_groups:
  - label: DynamoDB Stream & Billing
    variables:
      - billing_mode
      - stream_enabled
  - label: Advanced Settings
    variables:
      - server_side_encryption

The above configuration will:

• Show Billing Mode as a required dropdown.
• Render encryption as a form with Enabled and KMS Key ARN fields.
• Split variables into DynamoDB Stream & Billing and Advanced Settings sections.

Click to view variable fields

Field	Description
label	Display name in the config panel
description	Help text
type	Terraform type (string, bool, object, list, and so on)
terraform_variable_type	Explicit Terraform type when it differs from type
ui_control	UI widget (text, dropdown, select, object, and so on)
options	Allowed values for dropdown or select
value_type	Value type hint for the UI
validation	Rules such as required or maxObjects
attributes	Nested fields for object or list items
default_editor	Editor for complex types (for example form)
variable_label	Label for a nested object block

ui_control values: text, password, select, dropdown, dictionary, textarea, object, hcl, yaml, json, shell, markdown, text_list, textarea_list, dictionary_list, object_list.

Terraform variable defaults still come from variables.tf, not from stackgen.yaml.

any and any(list) variables

Previously, complex open-ended types did not have a reliable attributes UI. Now, variables of Terraform type any or any(list) open in HCL Expression (Terraform expression) mode by default.

• The attributes UI stays in expression mode for these types. There is no deterministic form UI for arbitrary any values, so you do not switch to a normal editor for that field.
• Override that default only by explicitly defining a ui_control (or ui-control) in stackgen.yaml.
• If an any or any(list) variable has no default value, StackGen flags it as Action Required so incomplete open-ended inputs show up before you move on.

See v2026.7.9.

Connections

The connections list tells StackGen how to wire modules together on the topology. Each entry has a source module, a target module, an optional label (for example DATA or TRIGGER), and references that map outputs to inputs.

Example

connections:
  - source: aws_s3_bucket
    target: aws_lambda_function
    label: DATA
    references:
      - from: aws_s3_bucket
        to: aws_lambda_function
        output: arn
        input: bucket_arn
  - source: aws_sqs_queue
    target: aws_lambda_function
    label: TRIGGER
    references:
      - from: aws_sqs_queue
        to: aws_lambda_function
        output: queue_url
        input: queue_url

The above configuration will:

• Connect S3 to Lambda with a DATA link (arn to bucket_arn).
• Connect SQS to Lambda with a TRIGGER link (queue_url to queue_url).

Click to view connection fields

Field	Description
source	Source module name
target	Target module name
label	Connection type (DATA, TRIGGER, and so on)
references	Output-to-input mappings
references[].from	Module that exposes the output
references[].to	Module that receives the input
references[].output	Output attribute name
references[].input	Input attribute name

Discovery

The discovery block tells StackGen how to import an existing resource. Set the Terraform resource type, the data_source to read it, whether drag-and-drop discovery is allowed (dnd_supported), and a references map from stackgen attributes to resource outputs.

Example

discovery:
  type: aws_dynamodb_table
  dnd_supported: false
  data_source: data_aws_dynamodb_table
  references:
    table_name:
      from: aws_dynamodb_table
      output: name
    table_arn:
      from: aws_dynamodb_table
      output: arn

The above configuration will:

• Discover aws_dynamodb_table resources using data_aws_dynamodb_table.
• Map table_name and table_arn from the resource outputs.
• Disable drag-and-drop discovery (dnd_supported: false).

Click to view discovery fields

Field	Description
type	Terraform resource type to discover
dnd_supported	true or false for drag-and-drop discovery
data_source	Terraform data source name
references.<name>.from	Resource the output comes from
references.<name>.output	Output attribute name

Click to view full schema (v1)

version: string

representation:
  description: string              # optional
  icon: string                     # optional
  side_panel:
    label: string
    icon: string
  node:
    label:
      static: string
      template: string
    display_type: string
    label_attribute: string        # optional

container:                         # optional
  type: string                     # optional
  on_delete: string                # optional; "cascade" | "detach"
  connection_rules:                # optional
    - source_module_name: string
      connection_type: string      # optional
      mapping:
        source_attribute: string
        target_attribute: string
        data_type: string          # optional

variables:
  <name>: string                   # map key
    label: string                  # optional
    description: string            # optional
    ui_control: string             # optional
    options: any[]                 # optional
    value_type: string             # optional
    validation:                    # optional
      <key>: any
    attributes:                    # optional; recursive Variable map
      <name>: ...                  # same Variable shape
    default_editor: string         # optional
    variable_label: string         # optional
    type: string                   # optional
    terraform_variable_type: string # optional

variable_groups:
  - label: string
    description: string
    variables: string[]

connections:
  - source: string
    target: string
    label: string                  # optional
    references:                    # optional
      - from: string
        to: string
        output: string
        input: string

discovery:
  type: string
  dnd_supported: boolean
  data_source: string
  references:
    <attribute_name>:
      from: string
      output: string

Click to view full DynamoDB example

version: "1.0"

representation:
  description: |
    DynamoDB table module with container, discovery, and connection metadata.
  icon: aws-dynamodb-table
  side_panel:
    label: DynamoDB Table
    icon: aws-dynamodb-table
  node:
    label:
      static: DynamoDB Table
      template: DynamoDB Table - ${name}
    display_type: container
    label_attribute: name

container:
  type: networking
  on_delete: cascade
  connection_rules:
    - source_module_name: aws_vpc
      connection_type: DATA
      mapping:
        source_attribute: vpc_id
        target_attribute: vpc_id
        data_type: string

variables:
  billing_mode:
    label: Billing Mode
    description: The billing mode of the DynamoDB table.
    ui_control: dropdown
    options:
      - PROVISIONED
      - PAY_PER_REQUEST
    value_type: string
    validation:
      required: true
    type: string
    terraform_variable_type: string

  stream_enabled:
    label: Stream Enabled
    description: Enable DynamoDB Streams.
    type: bool
    options:
      - true
      - false

  server_side_encryption:
    label: Server-Side Encryption
    description: Encryption settings for the table.
    type: object
    default_editor: form
    variable_label: Encryption block
    validation:
      required: false
      maxObjects: 1
    attributes:
      enabled:
        label: Enabled
        description: Whether server-side encryption is enabled.
        type: bool
      kms_key_arn:
        label: KMS Key ARN
        description: The ARN of the KMS key used for encryption.
        type: string

  global_secondary_index:
    type: list
    validation:
      maxObjects: 20
    attributes:
      name:
        label: Index Name
        type: string
      hash_key:
        label: Hash Key
        type: string
      projection_type:
        label: Projection Type
        type: string
        options:
          - ALL
          - KEYS_ONLY
          - INCLUDE

variable_groups:
  - label: DynamoDB Stream & Billing
    description: Configure DynamoDB stream and billing options.
    variables:
      - billing_mode
      - stream_enabled
  - label: Indexes
    description: Configure secondary indexes.
    variables:
      - global_secondary_index
  - label: Advanced Settings
    description: Advanced DynamoDB settings.
    variables:
      - server_side_encryption

connections:
  - source: aws_s3_bucket
    target: aws_lambda_function
    label: DATA
    references:
      - from: aws_s3_bucket
        to: aws_lambda_function
        output: arn
        input: bucket_arn
      - from: aws_s3_bucket
        to: aws_lambda_function
        output: id
        input: bucket_id
  - source: aws_sqs_queue
    target: aws_lambda_function
    label: TRIGGER
    references:
      - from: aws_sqs_queue
        to: aws_lambda_function
        output: queue_url
        input: queue_url

discovery:
  type: aws_dynamodb_table
  dnd_supported: false
  data_source: data_aws_dynamodb_table
  references:
    table_name:
      from: aws_dynamodb_table
      output: name
    table_arn:
      from: aws_dynamodb_table
      output: arn