AI Shell

To skip the explanation section and only receive the command, use the -s or --silent flag. You can also make this a permanent preference.

One-time use:

ai -s 列出所有日志文件

Set as permanent preference:

ai config set SILENT_MODE=true

Install the ai-shell CLI globally using npm. Requires Node.js v14 or higher.

1. Install the package:

   npm install -g @builder.io/ai-shell

2. Obtain an API key from OpenAI.
3. Configure your key:

   ai config set OPENAI_KEY=<your token>

   This creates a .ai-shell configuration file in your home directory.

To convert a natural language prompt into a shell command, use the ai command followed by your prompt.

Basic Usage:

ai <prompt>

Example:

ai list all log files

After running a prompt, the CLI will provide the suggested script and an explanation. You can then choose to:

• Yes: Run the script.
• Revise: Edit the command via a new prompt.
• Cancel: Abort the operation.

Tip: If your prompt contains special characters like ? or * that your shell might interpret incorrectly, wrap the prompt in quotes:

ai 'what is my ip address'

You can manage configuration via the CLI or an interactive UI.

Interactive UI: Run the following command to open a visual interface to set options like OpenAI Key, API Endpoint, Silent Mode, Model, and Language:

ai config

Manual Configuration via CLI: Use ai config set <KEY>=<VALUE> to update specific settings.

AI Shell is a CLI that converts natural language to shell commands. It requires Node.js v14 or higher.

1. Install the package globally via npm:

   npm install -g @builder.io/ai-shell

2. Obtain an OpenAI API key from OpenAI.
3. Configure the key in AI Shell:

   ai config set OPENAI_KEY=<your token>

   This creates a .ai-shell file in your home directory.

To check your current version:

ai --version

To upgrade to the latest version, you can either use npm:

npm update -g @builder.io/ai-shell

Or use the built-in update command:

ai update

To disable the explanation section in the output, you can use the --silent flag for a single command or set it as a permanent preference.

One-time use:

ai -s <prompt>
# or
ai --silent <prompt>

Permanent setting:

ai config set SILENT_MODE=true

ai -s list all log files

You can manage your configuration using the ai config command.

Interactive UI

To open an interactive visual interface to view and set options (like OpenAI Key, Model, or Language), run:

ai config

Command Line Configuration

You can set specific keys directly using ai config set <KEY>=<VALUE>. Supported configuration keys include:

• OPENAI_KEY: Your OpenAI API token.
• SILENT_MODE: Set to true to skip command explanations.
• OPENAI_API_ENDPOINT: Custom OpenAI API endpoint (defaults to https://api.openai.com/v1).
• LANGUAGE: The language for AI responses (e.g., en, zh-Hans, es).

AI Shell provides an interactive menu to view and update your settings. This UI allows you to:

• Enter or update your OPENAI_KEY.
• Change the OPENAI_API_ENDPOINT.
• Toggle SILENT_MODE on or off.
• Select a model from a list of available models fetched from your API endpoint.
• Change the interface LANGUAGE.
• View current values (with the API key obfuscated).

If you encounter a 429 error, it indicates you have exceeded your OpenAI quota or have incorrect billing settings.

To fix this:

1. Check your usage and billing at the OpenAI Billing Overview.
2. Ensure you have an active payment method added if you have exhausted any free credits.

If you encounter a 429 error, it typically indicates that you have exceeded your OpenAI quota or have an incorrect billing setup.

To resolve this:

1. Check your usage and billing status at OpenAI Billing.
2. Ensure a valid payment method is attached if you are not using an active grant.