quickx

quickx is a Codex profile manager. It lets you store multiple provider configurations and switch between them instantly — no manual editing of ~/.codex/config.toml required.

Getting Started

npm install -g @amethyst-labs/quickx

The fastest way to get going is via a provider template:

# Browse built-in and community templates
quickx templates list

# Preview what a template needs
quickx templates preview openai

# Create a profile from a template (prompts for API key, model, etc.)
quickx profiles add my-openai --from-template openai

# Activate it
quickx use my-openai

Or log in with your ChatGPT / Codex account:

quickx profiles login            # opens browser
quickx profiles login --device   # device-code flow (no browser)

CLI

Run with no arguments to open the interactive TUI.

Profile management

profiles is the primary command. config is a supported alias.

quickx profiles list
quickx profiles add <name> [options]
quickx profiles edit <name> [options]
quickx profiles remove <name>
quickx profiles login [name] [--device]

add options

Flag	Description
--from-template <id>	Bootstrap fields from a provider template
--base-url <url>	Provider API base URL
--api-key <key>	API key
--model <model>	Default model
--wire-api <api>	responses (default) or chat
--auth-method <method>	api_key (default) or chatgpt
--reasoning-effort <level>	Codex reasoning effort
--model-verbosity <level>	Codex model verbosity

edit accepts the same flags; only the flags you supply are changed.

Activation & status

quickx use <name>    # write profile to ~/.codex/config.toml
quickx status        # show active profile and all saved profiles

Templates

quickx templates list              # list all available templates
quickx templates preview <id>      # show fields and dynamic placeholders

Templates are fetched from AmethystDev-Labs/QuickCLI and merged with built-in presets. Remote templates are cached for 1 hour in ~/.cache/quickx/template-cache/.

TUI

Running quickx with no arguments launches a full-screen terminal UI built with Ink (React for the terminal).

Tabs — switch with 1 / 2 / 3

Tab	Contents
Status	Active profile, file paths, all saved profiles at a glance
Profiles	Scrollable profile list with a detail panel
Templates	Scrollable template list with a live preview panel

Profiles tab keys

Key	Action
↑ / ↓	Move selection
Enter or Ctrl+U	Activate selected profile
Ctrl+A	Open Add Profile form
Ctrl+E	Open Edit Profile form
Ctrl+D	Confirm-delete selected profile
Ctrl+L	Open Codex Login form
Ctrl+R	Refresh

Templates tab keys

Key	Action
↑ / ↓	Move selection
Enter	Open Create Profile form for selected template
Ctrl+R	Refresh

Global keys: Ctrl+Q or Esc to quit.

All forms use ↑ / ↓ to move between fields, printable characters to type, Backspace to delete, and Ctrl+S to submit. Esc cancels and returns to browse mode.

Profile fields

Field	Description
name	Internal profile key
displayName	Human-readable label
baseUrl	Provider API base URL
apiKey	API key (api_key auth mode)
model	Default model
wireApi	responses or chat
authMethod	api_key or chatgpt
reasoningEffort	Codex reasoning level
modelVerbosity	Optional Codex verbosity override

Storage

Path	Purpose
~/.config/quickx/config.json	quickx profile store (macOS/Linux)
%APPDATA%\quickx\config.json	quickx profile store (Windows)
~/.cache/quickx/template-cache/	Remote template cache (1 h TTL)
~/.codex/config.toml	Written on every quickx use
~/.codex/auth.json	Updated for ChatGPT auth mode

Release

Publishing is tag-driven. Push a v* tag and GitHub Actions publishes to npm.