Add support for configs

defang/POWER.md

@@ -2,6 +2,7 @@
 name: "defang"
 displayName: "Defang"
 description: "Deploy Docker Compose apps to your own cloud account using Defang."
+author: "DefangLabs"
 keywords:
   [
     "deploy",
@@ -31,9 +32,9 @@ Ensure that the step above ## Validate tools is completed before proceeding to t
 
 - Ask the user which cloud provider they want to use (AWS, GCP)
   - If AWS:
-    - Use the steering file `getting-started-aws`
+    - Use the steering file `deploying-to-aws`
   - If GCP:
-    - Use the steering file `getting-started-gcp`
+    - Use the steering file `deploying-to-gcp`
 
 # Available Steering Files
 

defang/steering/deploying-to-aws.md

@@ -14,7 +14,7 @@ When the user's project has a compose.yaml file and would like to deploy to aws.
 
 ## Step 1: Select the stack to which the application will be deployed
 
-A stack is a single deployed instance of your project in a specific AWS account and region. You can deploy multiple stacks into the same account and region, for example: `staging` and `production`.
+A [stack](https://docs.defang.io/docs/concepts/stacks) is a single deployed instance of your project in a specific AWS account and region. You can deploy multiple stacks into the same account and region, for example: `staging` and `production`.
 
 Check if there are any stacks in the current project.
 
@@ -25,7 +25,13 @@ The following information will be needed to create a stack:
 
 - Stack name: must be alphanumeric and must not start with a number
 - Region: for example: `us-west-2`
-- AWS Profile: the AWS profile with which the user should authenticate to AWS
+- AWS Profile: the AWS profile with which the user should authenticate to AWS.
+  - First, verify the AWS CLI is installed and configured by running `aws`.
+  - List available profiles with `aws configure list-profiles` and prompt the user to select one.
+  - If the AWS CLI is not installed or configured:
+    - Direct the user to the [installation guide](https://docs.aws.amazon.com/cli/latest/userguide/getting-started-install.html).
+    - Once installed, guide the user to configure an AWS Profile by running `aws configure` and entering their credentials.
+    - Restart the stack creation process.
 - Deployment Mode: The deployment mode is the primary parameter for managing the cost and resiliency of your application's deployment. The following deployment modes are available: `affordable`, `balanced`, and `high_availability`. The default is `affordable`. Learn more at https://docs.defang.io/docs/concepts/deployment-modes
 
 If a new stack is created, make sure to select it before it can be used.
@@ -34,6 +40,10 @@ If a new stack is created, make sure to select it before it can be used.
 
 Now that a stack is selected, the project can be deployed.
 
+### Configs
+
+The deployment will fail if any of the config values required by the project are not set. Please refer to the steering file `managing-configs` for more information on how to manage configs.
+
 ## Step 3: Monitor the deployment
 
 Once the deployment has begun, progress can be monitored by tailing the logs or periodically checking service status.

defang/steering/deploying-to-gcp.md

@@ -10,16 +10,16 @@ This guide describes how to deploy applications to GCP with Defang.
 
 ## Trigger
 
-When the user's project has a compose.yaml file and would like to deploy to aws.
+When the user's project has a compose.yaml file and would like to deploy to gcp.
 
 ## Step 1: Select the stack to which the application will be deployed
 
-A stack is a single deployed instance of your project in a specific AWS account and region. You can deploy multiple stacks into the same account and region, for example: `staging` and `production`.
+A [stack](https://docs.defang.io/docs/concepts/stacks) is a single deployed instance of your project in a specific GCP project and region. You can deploy multiple stacks into the same project and region, for example: `staging` and `production`.
 
 Check if there are any stacks in the current project.
 
 - If a stack is already defined, ask the user if they would like to select one of the existing stacks, or if they would like to create a new one.
-- If there are no stacks, prompt user to create a new AWS stack.
+- If there are no stacks, prompt user to create a new GCP stack.
 
 The following information will be needed to create a stack:
 
@@ -34,6 +34,10 @@ If a new stack is created, make sure to select it before it can be used.
 
 Now that a stack is selected, the project can be deployed.
 
+### Configs
+
+The deployment will fail if any of the config values required by the project are not set. Please refer to the steering file `managing-configs` for more information on how to manage configs.
+
 ## Step 3: Monitor the deployment
 
 Once the deployment has begun, progress can be monitored by tailing the logs or periodically checking service status.

defang/steering/managing-configs.md

@@ -0,0 +1,78 @@
+# Managing Configs
+
+Configs are key-value pairs that can be used to store sensitive information such as API keys, database credentials, or any other configuration data that should not be hardcoded into your application.
+
+**IMPORTANT**:
+Do not assume default values for configs; always prompt the user for input. For example, do not assume a default password for a database config is always randomly generated. Always present the user with the options to provide a value or choose to generate a random one before setting each config.
+
+## Viewing Configs
+
+To view the current configs for your project, use the following MCP tool:
+
+```bash
+list_configs
+```
+
+This will display a list of all the configs currently set for your project. This list does not represent configs you are still required to set; it only shows what has already been set.
+
+## Setting Configs
+
+To set a config, use the following MCP tool:
+
+```bash
+set_config
+```
+
+When setting configs, make sure to ask the user for either a specific value or whether to generate a random value. Do not assume choices on behalf of the user.
+
+**IMPORTANT**:
+When using the `set_config` tool, ensure that only one of the following options is provided: either the `value` parameter or the `random` flag.
+Providing both will result in an error.
+Example:
+
+```json
+{
+  "name": "POSTGRES_PASSWORD",
+  "value": "helloworld123",
+  "random": true,
+  "working_directory": "."
+}
+```
+
+Available parameters:
+
+- `name` (required): The key for the config you want to set.
+- `value` (optional): The value for the config. Do not provide this parameter if you are using the `random` parameter.
+  Example:
+  ```json
+  {
+    "name": "POSTGRES_PASSWORD",
+    "value": "helloworld123",
+    "working_directory": "."
+  }
+  ```
+- `random` (optional): If this flag is provided, a random value will be generated for the config. Do not provide the `value` parameter if you are using this parameter.
+  Example:
+  ```json
+  {
+    "name": "POSTGRES_PASSWORD",
+    "random": true,
+    "working_directory": "."
+  }
+  ```
+
+## Deleting Configs
+
+To delete a config, use the following MCP tool:
+
+```bash
+remove_config
+```
+
+This will remove the specified config from your project.
+
+## Best Practices
+
+- Avoid hardcoding sensitive information in your codebase. Use configs instead.
+- Regularly rotate sensitive configs such as API keys and passwords.
+- Use descriptive keys for your configs to make them easily identifiable.