Skip to main content
The qckfx CLI provides a powerful command-line interface for running AI agents, validating configurations, and managing sessions. This reference covers all available commands and options.

​
Installation

The CLI is included with the @qckfx/agent package: 
npm install -g @qckfx/agent
Or use it directly with npx: 
npx @qckfx/agent "Your prompt here"

​
Quick Start

Get started quickly by initializing a project with default agents: 
# Initialize project with default agent configurations
qckfx init
Then use the CLI with direct prompts: 
# Basic usage
qckfx "What files are in this directory?"

# With a specific model
qckfx -m claude-sonnet-4 "Analyze this codebase"

# Using an agent configuration
qckfx -a my-agent.json "List all TODOs in the code"

​
Commands Overview

The qckfx CLI provides several key functionalities:
  • Project Initialization - Set up new projects with default agent configurations
  • Agent Execution - Run agents with prompts and configurations
  • Configuration Validation - Validate agent configuration files
  • Session Management - Continue previous conversations
  • Sub-agent Integration - Add sub-agents dynamically

Interactive Mode

If you run qckfx without a prompt, it will enter interactive mode and ask you to enter your prompt.

​
Environment Variables

The CLI respects several environment variables:
VariableDescriptionDefault
LLM_API_KEYAPI key for your model providerRequired
LLM_BASE_URLBase URL for your model providerProvider default
You can override these environment variables using the --api-key and --url command-line options.

​
Next Steps

⌘I