Shell-AI (TypeScript)

A CLI utility that uses AI to suggest shell commands from natural language descriptions. TypeScript port using OpenRouter as the LLM provider.

Based on shell-ai by Rick Lamers.

Installation

# Run directly with npx (no install needed)
npx @petah/shell-ai list all files

# Or install globally
npm install -g @petah/shell-ai

Usage

# Basic usage
shai list all docker containers

# With context mode (keeps command outputs for better suggestions)
shai --ctx show me running processes

Configuration

Shell-AI can be configured through environment variables or a config file.

Config File

Create a config file at:

• Linux/macOS: ~/.config/shell-ai/config.json
• Windows: %APPDATA%\shell-ai\config.json

{
  "OPENROUTER_API_KEY": "your_openrouter_api_key",
  "OPENROUTER_MODEL": "anthropic/claude-3.5-sonnet",
  "SHAI_SUGGESTION_COUNT": 3,
  "SHAI_TEMPERATURE": 0.05,
  "CTX": false
}

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | OPENROUTER_API_KEY | Your OpenRouter API key (required) | - | | OPENROUTER_MODEL | Model to use | anthropic/claude-3.5-sonnet | | SHAI_SUGGESTION_COUNT | Number of suggestions | 3 | | SHAI_SKIP_CONFIRM | Skip command confirmation | false | | SHAI_SKIP_HISTORY | Skip writing to shell history | false | | SHAI_TEMPERATURE | LLM temperature (0-1) | 0.05 | | CTX | Enable context mode | false | | DEBUG | Enable debug output | false |

Available Models

OpenRouter supports many models. Some popular options:

• anthropic/claude-3.5-sonnet (default)
• openai/gpt-4-turbo
• openai/gpt-3.5-turbo
• google/gemini-pro
• meta-llama/llama-3-70b-instruct

See OpenRouter Models for the full list.

Features

• Natural language to shell commands
• Dynamic command detection - automatically detects available commands on your system
• Interactive command selection
• Command confirmation before execution
• Shell history integration (zsh, bash, fish, etc.)
• Context mode for multi-step workflows
• Cross-platform support (Linux, macOS, Windows)

License

MIT