Configuration Overview

Every Watt TF project centers around two structural layers: the Input Data (what your applications or developers require) and the Blueprint Configuration (how those requirements map to infrastructure).

This section covers the basic syntax, file structure, and top-level fields available in a Watt TF blueprint.

The Anatomy of a Blueprint

A blueprint is a standard YAML file. Its primary purpose is to hold an array of declarative instructions inside the transform key.

Here is a minimalist but complete blueprint structure:

# blueprint.yaml

# The transform block is the core array containing all orchestration steps
transform:
  - target: resource.local_file.welcome
    value:
      filename: "welcome.txt"
      content: "Hello World"

Inside a Transform Block

Each object inside the transform array is a separate operation evaluated in strict sequential order. A transform block accepts the following configuration keys:

transform:
  - if: <CEL-Expression>       # (Optional) Conditional execution check
    for_each: <CEL-Expression> # (Optional) Collection to iterate over
    target: <String-Path>      # (Required) The dot-notated destination path
    value: <Any-Type>          # (Required) The data structure or primitive to inject

• target (The Destination): Points to the exact location in the final Terraform configuration where the value should live (e.g., resource.aws_s3_bucket.main).
• value (The Payload): Can be a single string, an integer, a boolean, an array, or a deeply nested map/object. Watt TF evaluates all dynamic variables inside this payload before placement.
• if (The Gatekeeper): A common expression language (CEL) string returning a boolean. If it yields false, the execution of this entire block terminates instantly.
• for_each (The Multiplier): A CEL expression that evaluates to a list or a range. It causes the block to clone itself and execute once for each element in the collection.

Next Steps

Now that you know how a blueprint is structured globally, dive deeper into how the individual engines behave:

• Transformations & Value Mapping - Master how data is injected into payloads.
• Path Resolution & Targets - Learn how dot-notation automatically builds nested JSON.
• Variable Interpolation - Read values from inputs and environment variables effortlessly.