Project Configuration
Reference for the gct.yaml configuration file and environment variables.
Configuration Reference
GCT offers a flexible configuration system that allows you to set up your AI provider and other settings through files and environment variables.
Configuration Hierarchy
GCT loads configuration from the following sources in order, with later sources overriding earlier ones:
- Global Config File: A
config.yamlfile located in your user config directory. This is useful for setting a default provider and API key that you use across all your projects.- Linux/macOS:
~/.config/gct/config.yaml(or~/.gct/config.yamlfor legacy) - Windows:
%APPDATA%\gct\config.yaml
- Linux/macOS:
- Local Config File: A
gct.yamlfile in your project's directory (or any parent directory). This allows you to have project-specific settings, like custom commit guidelines. - Environment Variables: Any variable prefixed with
GCT_. These are perfect for CI/CD environments or for temporarily overriding a setting without modifying a file. You can also place these in a.envfile in your project directory for them to be loaded automatically.
gct.yaml File Reference
The gct.yaml file is the heart of GCT's AI capabilities. The gct init and gct init model commands will generate this file for you automatically.
Top-Level Fields
| Field | Type | Required | Description |
|---|---|---|---|
name | string | No | A friendly name for your project. |
provider | string | Yes | The AI provider to use (e.g. OpenAI, Anthropic, Google AI Studio). Must match one of the Supported Providers. |
model | string | Yes | The specific model or deployment ID for the chosen provider (e.g. gpt-4o, claude-3-5-haiku-latest). |
api | string | Yes | Your secret API key for the chosen provider. This is a secret and should not be committed. |
cache | object | No | Settings for caching AI responses to reduce costs and latency. |
Provider-Specific Fields
These fields are only required if you are using the specified provider.
| Field | Provider | Required | Description |
|---|---|---|---|
endpoint | OpenAI Compatible | Yes | The base URL of the API endpoint (e.g. https://api.example.com/v1). |
gcp_project_id | Google Vertex AI | Yes | Your Google Cloud Platform Project ID. |
gcp_region | Google Vertex AI | Yes | The GCP region for your Vertex AI model (e.g. us-central1). |
aws_region | Amazon Bedrock | Yes | The AWS region where your Bedrock models are hosted (e.g. us-east-1). |
aws_access_key_id | Amazon Bedrock | Yes | Your AWS Access Key ID for authentication. |
aws_secret_access_key | Amazon Bedrock | Yes | Your AWS Secret Access Key for authentication. |
azure_resource_name | Azure OpenAI | Yes | The name of your Azure OpenAI resource. |
Custom Guidelines
| Field | Type | Required | Description |
|---|---|---|---|
commits.guides | array | No | A list of paths to local .md or .txt files that will be used as guidelines for gct ai commit. |
changelogs.guides | array | No | A list of paths to local .md or .txt files that will be used as guidelines for gct ai log. |
Environment Variables
All configuration fields can be set using environment variables. This is especially useful in CI/CD environments.
| Environment Variable | gct.yaml Field | Required |
|---|---|---|
GCT_NAME | name | No |
GCT_PROVIDER | provider | Yes |
GCT_MODEL | model | Yes |
GCT_API_KEY | api | Yes |
GCT_ENDPOINT | endpoint | Only for OpenAI Compatible provider |
GCT_GCP_PROJECT_ID | gcp_project_id | Only for Google Vertex AI provider |
GCT_GCP_REGION | gcp_region | Only for Google Vertex AI provider |
GCT_AWS_REGION | aws_region | Only for Amazon Bedrock provider |
GCT_AWS_ACCESS_KEY_ID | aws_access_key_id | Only for Amazon Bedrock provider |
GCT_AWS_SECRET_ACCESS_KEY | aws_secret_access_key | Only for Amazon Bedrock provider |
GCT_AZURE_RESOURCE_NAME | azure_resource_name | Only for Azure OpenAI provider |
GCT_CACHE_ENABLED | cache.enabled | No |
Last updated on