A modern, feature-rich terminal UI chat application with OpenAI integration, built with OpenTUI and React.

👉 New here? Start with the Quick Start Guide | ✋ Read our Code of Conduct before contributing

• 🤖 OpenAI/Azure Integration - Streaming responses with OpenAI and Azure OpenAI support
• 🔌 LiteLLM-Compatible Providers - Use openai/... style model IDs via LiteLLM proxies
• 💬 Session Management - Multiple conversations with persistent history
• ⚡ Command System - 10+ built-in commands + custom command support
• 🏷️ Smart Mentions - Context, file, code, and docs references
• 🎯 Autocomplete - Fuzzy suggestions with keyboard navigation
• 🎨 Themes - Light/Dark toggle
• 💾 Settings Persistence - Preferences saved across sessions

• 🎨 Claude Code-Inspired Design - Clean, professional interface with warm accents
• 📱 Fully Responsive - Adapts to any terminal size
• ⌨️ Keyboard-Driven - Efficient workflows without leaving the keyboard
• 📜 Smart Scrolling - Auto-scroll with visual feedback
• 🔧 Overlay Menus - Settings and session list overlays

/clear, /help, /model, /endpoint, /api-key, /status, /settings, /sessions, /commands, /export, /theme, /terminal-setup, and more

Install from npm:

npm install -g qlaw-cli

Or using other package managers:

# Using yarn
yarn global add qlaw-cli

# Using pnpm
pnpm add -g qlaw-cli

# Using bun
bun add -g qlaw-cli

Then run from anywhere:

qlaw

If you want to contribute or develop locally:

# Clone the repository
git clone https://github.com/Qredence/qlaw-cli.git
cd qlaw-cli

# Install dependencies
bun install

# Copy environment template
cp .env.example .env

# Add your OpenAI API key to .env
# OPENAI_API_KEY=your-key-here

# LiteLLM (default provider) example
# LITELLM_BASE_URL=http://localhost:4000/v1
# LITELLM_API_KEY=your-key-here
# LITELLM_MODEL=openai/gpt-4o-mini
# LITELLM_MODELS=openai/gpt-4o-mini,openai/gpt-4o

# Run locally
bun run start

# Or with auto-reload during development
bun run dev

1. Type your message in the input field
2. Press Enter to send
3. AI responds with streaming support
4. Use commands by typing / for quick actions
5. Add mentions by typing @ for context

Type / to see available commands:

• /clear - Clear chat history (with confirmation)
• /help - Show help information
• /provider - Set provider (openai/azure/litellm/custom)
• /model - Set the model name
• /endpoint - Set the API endpoint base URL
• /api-key - Set the API key (masked in status)
• /tools - Toggle tool execution (read/list/write/run)
• /tools perm - Set tool permissions (allow/ask/deny)
• /status - Show current configuration
• /settings - Print current settings summary
• /settings panel - Open the interactive settings menu
• /sessions - View recent sessions
• /commands - List custom commands
• /export - Export current chat to JSON
• /theme - Toggle light/dark theme
• /terminal-setup - Terminal keybinding tips
• /keybindings - Inspect or edit suggestion navigation shortcuts
• /mode - Switch between standard/workflow modes
• /workflow - Workflow controls reference
• /agents - Show current agent-fleet roles
• /run - Kick off the workflow in the active mode
• /continue - Continue a workflow handoff
• /judge - Invoke the judge agent for a decision
• /af-bridge - Configure the Agent Framework bridge base URL
• /af-model - Configure the Agent Framework model identifier

• Run /settings panel to open the interactive settings overlay (Core API, UI, Agent Framework sections)
• Run /settings alone to print the current configuration in the transcript
• Use ↑ ↓ or Tab to highlight a row, Enter to edit/toggle, and Esc to close
• Text fields launch inline prompts; updates persist to ~/.qlaw-cli/qlaw_settings.json
• Agent Framework rows mirror /af-bridge + /af-model, and workflow mode can stay enabled by default
• Update suggestion navigation shortcuts with /keybindings set <action> <binding> or /keybindings reset

Type @ for contextual references:

• @context <text> - Add contextual information to your message
• @file <path> - Reference a file in your message
• @code <snippet> - Include a code snippet in your message
• @docs <topic> - Reference documentation in your message
• @code function example() { return true; } will format as a code snippet

Mentions are automatically formatted to provide structured context to the AI. For example:

• @docs API authentication will format as a documentation reference
• @file src/index.ts will inline the file contents (truncated if needed)

When /tools is enabled, the assistant can request tool execution using fenced tool blocks. Permissions follow allow | ask | deny and can be configured via /tools perm. run_command executes shell commands and should only be enabled when you trust the model.

• ↑ ↓ - Navigate suggestions
• Tab - Autocomplete suggestion
• Enter - Send message / Select suggestion
• Esc - Cancel input / Close overlays / Exit
• Ctrl+C - Force exit

• Quick Start Guide - Get up and running in 3 minutes
• Architecture - Technical design and structure
• UI Reference - Visual interface guide
• Design System - Colors, typography, and components
• Changelog - Version history
• API Integration - Configure OpenAI/Azure/custom backends
• Agent Bridge Example - Python Agent Framework bridge used by AF modes

The next release will focus on the following key areas:

• Improved command autocomplete and suggestions
• Interactive prompts and confirmations
• Enhanced keyboard navigation and shortcuts
• Real-time feedback and visual indicators

• Comprehensive settings panel
• User preferences persistence
• Configurable themes and colors
• API configuration management
• Custom keybindings

• Native integration of agent-framework as core framework
• Native integration of agentic-fleet as core framework
• Seamless agent orchestration and management
• Multi-agent conversation support
• Agent capability discovery

• Streamlined onboarding flow
• Enhanced error messages and help system
• Performance optimizations
• Accessibility enhancements
• Improved session management UI

• Multi-model support (Claude, Gemini, etc.)
• Local LLM integration (Ollama)
• Voice input support
• Image analysis capabilities
• Advanced RAG with vector search
• Plugin marketplace
• Collaborative sessions
• Cloud sync capabilities

We welcome contributions! Please see:

• Contributing Guide - How to contribute
• Code of Conduct - Community standards
• Security Policy - Reporting vulnerabilities
• Publishing Guide - For maintainers: How to publish releases to npm

MIT License - see LICENSE file for details.

Inspiration:

• Claude Code - Design inspiration for the clean, minimal interface and warm accent color scheme
• Cursor - Terminal integration patterns and workflow concepts

Built With:

• OpenTUI - The excellent terminal UI framework that powers this application
• React - Component library for building the interface
• Bun - Fast JavaScript runtime and package manager
• TypeScript - Type-safe development

Special thanks to the OpenTUI team for creating such a powerful and elegant framework for building terminal UIs.

Made with ❤️ by Qredence