CLI Introduction

The llmist CLI brings the power of AI agents directly to your terminal. Run completions, interactive agents with tools, generate images and speech, and create powerful automation workflows - all from the command line.

Why Use the CLI?

Quick Completions

Get AI responses without writing any code - pipe input, get output

Interactive Agents

Run agents with gadgets (tools) in an interactive TUI

Automation

Build powerful pipelines with stdin/stdout integration

Custom Commands

Define reusable commands in TOML configuration

Key Features

• 8 built-in commands - complete, agent, image, speech, vision, models, gadget, init
• TUI mode - Beautiful terminal interface with expandable blocks and keyboard navigation
• Gadget ecosystem - Load gadgets from local files, npm packages, or git repositories
• TOML configuration - Define custom commands, templates, and workflows
• Gadget approval - Security controls for tool execution

Quick Example

bun

# Simple completion
bunx @llmist/cli complete "What is the capital of France?"

# Interactive agent (TUI activates automatically in terminals)
bunx @llmist/cli agent "Help me organize my project"

# Generate an image
bunx @llmist/cli image "A sunset over mountains" -o sunset.png

# Custom command from config
bunx @llmist/cli code-review ./src/main.ts

npm

# Simple completion
npx @llmist/cli complete "What is the capital of France?"

# Interactive agent (TUI activates automatically in terminals)
npx @llmist/cli agent "Help me organize my project"

# Generate an image
npx @llmist/cli image "A sunset over mountains" -o sunset.png

# Custom command from config
npx @llmist/cli code-review ./src/main.ts

Tip

The CLI automatically discovers API keys from environment variables (OPENAI_API_KEY, ANTHROPIC_API_KEY, GEMINI_API_KEY).

Next Steps

Installation Install the CLI globally

Quick Start Run your first agent

Commands Reference All available commands

Configuration TOML config and custom commands