Command Line Interface (CLI)

ABOV3's powerful command-line interface for automation, scripting, and headless operation.

Quick Start

# Show help and available commands
abov3 --help

# Start TUI (Terminal User Interface)
abov3

# Run a single prompt
abov3 run "Explain async/await in JavaScript"

# Continue last session
abov3 --continue

🔄 Self-Management Commands (New in v0.1.6)

ABOV3 includes built-in commands to manage its own installation. All functionality is compiled into the single binary - no external scripts needed!

abov3 install

Install ABOV3 to your system PATH automatically. Configures PATH based on your platform.

# Linux/macOS
./abov3 install

# Windows
.\abov3.exe install

# After installation, abov3 is available from anywhere
abov3 --version

✨ What this does:

Linux/macOS: Copies to ~/.local/bin/ and updates shell config
Windows: Copies to %LocalAppData%\Programs\ABOV3\ and updates registry PATH

abov3 update

Check for and install the latest version from GitHub releases. Performs atomic binary replacement with automatic rollback on failure.

# Check for updates and install
abov3 update

# Shows:
# - Current version
# - Latest available version
# - Changelog highlights
# - Download progress
# - Installation status

Example Output:

🔄 Checking for updates...

📦 Current version: 0.1.5
📦 Latest version:  0.1.6

📝 What's new:
──────────────────────────────────────────────────
- Self-management commands
- Ollama continuous monitoring
- Windows connection fixes

Proceed with update? (Y/n): y
⬇️  Downloading abov3.exe (143.00 MB)...
✓ Download complete
✓ Update installed

✨ Successfully updated to version 0.1.6

abov3 uninstall

Remove ABOV3 from your system, including binary and PATH configuration.

# Remove ABOV3
abov3 uninstall

# Cleans up:
# - Binary file
# - PATH configuration
# - Shell config entries (Linux/macOS)
# - Registry entries (Windows)

Main Commands

abov3 [project]

Start ABOV3 Terminal User Interface (TUI). This is the default command when no arguments are provided.

# Start TUI in current directory
abov3

# Start TUI in specific project directory
abov3 /path/to/project

abov3 run [message..]

Run ABOV3 with a single message or prompt. Useful for one-off queries or automation.

# Ask a question
abov3 run "What is the difference between let and const?"

# Request code generation
abov3 run "Create a REST API with Express.js"

# With specific model
abov3 run -m anthropic/claude-3-opus "Optimize this function"

abov3 auth

Manage authentication and credentials for AI providers.

# Login to a provider (supports Anthropic OAuth)
abov3 auth login

# List configured providers
abov3 auth list

# Logout from a provider
abov3 auth logout

💡 Anthropic OAuth

ABOV3 supports OAuth login for Anthropic Claude Pro/Max accounts. Run auth login to authenticate with your Claude account.

abov3 serve

Start a headless ABOV3 server. Useful for running ABOV3 as a backend service.

# Start server on default port
abov3 serve

# Start on specific port
abov3 serve --port 4096

# Specify hostname and port
abov3 serve --hostname 0.0.0.0 --port 8080

abov3 models

List all available AI models from configured providers.

# List all models
abov3 models

# Shows models from:
# - Anthropic (Claude models)
# - OpenAI (GPT models)
# - Ollama (local models)
# - Other configured providers

🏠 Ollama Continuous Monitoring (New in v0.1.6)

ABOV3 now continuously monitors Ollama every 30 seconds. Models automatically appear when Ollama starts and disappear when it stops. No manual refresh needed!

abov3 config

Manage ABOV3 configuration settings.

# Open configuration menu
./abov3-linux-x64 config

# Configuration is stored in:
# ~/.config/abov3/abov3.json

abov3 agent

Manage AI agents (Build, Plan, General, and custom agents).

# Manage agents
./abov3-linux-x64 agent

# Agents are configured in:
# ~/.config/abov3/agent/

abov3 export [sessionID]

Export session data as JSON for backup or analysis.

# Export current session
./abov3-linux-x64 export

# Export specific session
./abov3-linux-x64 export session_12345

abov3 upgrade [target]

Upgrade ABOV3 to the latest version or a specific version.

# Upgrade to latest
./abov3-linux-x64 upgrade

# Upgrade to specific version
./abov3-linux-x64 upgrade v1.2.3

Global Options

Option	Short	Description	Example
`--model`	`-m`	Specify the AI model to use	`-m anthropic/claude-3-opus`
`--continue`	`-c`	Continue the last session	`./abov3 -c`
`--session`	`-s`	Continue a specific session by ID	`-s session_123`
`--agent`		Specify which agent to use	`--agent build`
`--prompt`	`-p`	Provide initial prompt	`-p "Help me debug this"`
`--port`		Port for server mode	`--port 4096`
`--hostname`	`-h`	Hostname for server mode	`-h 127.0.0.1`
`--print-logs`		Print logs to stderr	`--print-logs`
`--log-level`		Set log level (DEBUG, INFO, WARN, ERROR)	`--log-level DEBUG`
`--version`	`-v`	Show version number	`./abov3 -v`
`--help`		Show help information	`./abov3 --help`

Common Workflows

Quick Code Generation

# Generate a Python function
./abov3-linux-x64 run "Create a Python function to parse CSV files"

# Generate with specific model
./abov3-linux-x64 run -m ollama/qwen2.5-coder:14b "Create a REST API"

Continuous Development Session

# Start a new session with a specific model
./abov3-linux-x64 -m anthropic/claude-3-sonnet

# Continue your last session
./abov3-linux-x64 --continue

# Continue a specific session
./abov3-linux-x64 --session session_20240315_142350

Using Local Models with Ollama

# First, ensure Ollama is running
ollama serve

# List available models (including Ollama)
./abov3-linux-x64 models

# Use a local Ollama model
./abov3-linux-x64 run -m ollama/mistral:7b "Explain closures"

Debugging and Troubleshooting

# Enable debug logging
./abov3-linux-x64 --log-level DEBUG --print-logs

# Check provider authentication
./abov3-linux-x64 auth list

# Test with a simple prompt
./abov3-linux-x64 run --print-logs "Hello, are you working?"

Environment Variables

# Set default model
export ABOV3_MODEL="anthropic/claude-3-opus"

# Set Ollama base URL (if not localhost)
export OLLAMA_BASE_URL="http://192.168.1.100:11434"

# Enable debug logging
export LOG_LEVEL="DEBUG"

# Set config directory
export ABOV3_CONFIG_DIR="$HOME/.config/abov3"

Configuration Files

ABOV3 uses the following configuration files:

~/.config/abov3/abov3.json - Main configuration
~/.config/abov3/agent/*.md - Agent configurations
~/.config/abov3/command/*.md - Custom commands
~/.local/share/abov3/auth.json - Authentication tokens
.abov3/ - Project-specific configuration

💡 Pro Tips

Use --continue to resume your last conversation without losing context
The TUI (abov3-tui) provides a better interactive experience than CLI
Combine with shell scripts for powerful automation workflows
Use --print-logs when debugging issues
Local Ollama models work offline and have no API costs