### 🔹 `inputs`
Defines the values this module requires as inputs from other modules.
#### ✅ Syntax:
```yaml
inputs:
<input_name>:
type: @outputs/<type>
optional: true|false
displayName: Display name for UI
description: Description of the input
```
#### 🔑 Common Fields:
- **`type`**: Required. Specifies the output type from another module (e.g. `@outputs/kubernetes-cluster`).
- **`optional`**: Boolean. Whether this input is required (default: `false`).
- **`displayName`**: String. Human-readable name shown in the UI.
- **`description`**: String. Explanation of what this input is used for.
#### 💡 Example:
```yaml
inputs:
cluster_config:
type: "@outputs/kubernetes-cluster"
optional: false
displayName: "Kubernetes Cluster"
description: "The target Kubernetes cluster for deployment"
storage_account:
type: "@outputs/azure-storage-account"
optional: true
displayName: "Storage Account (Optional)"
description: "Azure storage account for persistent data"
```
---
### 🔹 `outputs`
Defines the values this module exposes for consumption by other modules.
#### ✅ Syntax:
```yaml
outputs:
<output_name>:
type: @outputs/<type>
title: Will appear on the UI
```
#### 🔑 Common Fields:
- **`type`**: Required. Specifies the classification of the output (e.g. `@outputs/databricks-account`).
- **Use hyphens** (`-`) in the type name instead of underscores (`_`) if needed.
- **`output_attributes` and `output_interfaces` local variables**: These generate Terraform `output` blocks in **runtime**:
- `output_attributes` → corresponds to `output "attributes" { ... }`
- `output_interfaces` → corresponds to `output "interfaces" { ... }`
<important> Never generate output blocks for facets modules</important>
#### 💡 Special Notes:
- **`default`** is a **reserved keyword** that refers to the full output of the module. It is treated as the default
export and typically maps to the entire structured response from Terraform.
- A module can expose **multiple outputs**, including specific nested fields within the primary structure.
- Use dot notation to reference these nested fields explicitly:
```yaml
outputs:
default:
type: "@outputs/gcp-project"
title: "The GCP Project"
attributes.project_id:
type: "@outputs/project-id"
title: "The GCP Project id"
```
<important> no need to add properties in the outputs block like inputs.</important>
This allows consuming modules to wire only the specific part of the output they require, while still having the
option to consume the entire object via `default`.
---
### 🔐 Marking Output Fields as Sensitive
When calling `write_outputs`, every field requires both `value` and `sensitive` keys. The value can be any type (string, bool, number, list, dict).
```python
write_outputs(
module_path="/path/to/module",
output_attributes={
# String values
"instance_id": {"value": "aws_instance.main.id", "sensitive": False},
"api_key": {"value": "var.api_key", "sensitive": True},
# Complex types
"config": {"value": {"region": "us-east-1", "zones": ["a", "b"]}, "sensitive": False},
# Do not commit literal secrets to VCS; use variables or data sources
"credentials": {"value": ["data.aws_secretsmanager_secret.api_key.secret_string", "data.aws_secretsmanager_secret.db_password.secret_string"], "sensitive": True},
"enabled": {"value": True, "sensitive": False}
}
)
```
Generates:
```hcl
locals {
output_attributes = {
instance_id = aws_instance.main.id
api_key = sensitive(var.api_key)
config = {
region = "us-east-1"
zones = ["a", "b"]
}
credentials = sensitive(["token1", "token2"])
enabled = true
secrets = ["api_key", "credentials"] # Auto-generated for Facets UI
}
}
```
<important>Mark any field containing passwords, tokens, keys, or credentials as sensitive.</important>
---
