Install Matrix¶

Use this matrix to choose the cleanest deployment path per environment.

Runtime matrix¶

Platform	Backend runtime	WhatsApp bridge	Service mode	Notes
Linux	Native Python venv	Native Node.js 20+	systemd user units	Recommended primary path
macOS	Native Python venv	Native Node.js 20+	LaunchAgent or foreground	systemd user units are Linux-only
Windows	WSL2 (recommended)	WSL2 Node.js	user shell / task scheduler	Keep runtime in one WSL distro

Quick install commands¶

Linux/macOS¶

git clone https://github.com/galyarderlabs/galyarder-agent.git
cd galyarder-agent/backend/agent
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
g-agent onboard
g-agent status

Windows (WSL2)¶

git clone https://github.com/galyarderlabs/galyarder-agent.git
cd galyarder-agent/backend/agent
python3 -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
g-agent onboard
g-agent status

Script-based install/uninstall¶

If you prefer bootstrap scripts instead of manual setup, use deploy/*:

Arch Linux¶

Install: bash deploy/arch/install.sh
Uninstall: bash deploy/arch/uninstall.sh

Debian/Ubuntu¶

Install: bash deploy/debian/install.sh
Uninstall: bash deploy/debian/uninstall.sh

macOS¶

Install: bash deploy/macos/install.sh
Uninstall: bash deploy/macos/uninstall.sh

Windows (PowerShell)¶

Install: powershell -ExecutionPolicy Bypass -File deploy/windows/install.ps1
Uninstall: powershell -ExecutionPolicy Bypass -File deploy/windows/uninstall.ps1

Note: g-agent onboard is safe to re-run after upgrades — it merges new config defaults without overwriting your existing settings.

Post-install checklist¶

Confirm provider/model in ~/.g-agent/config.json
Set strict allowFrom for Telegram/WhatsApp
Keep tools.restrictToWorkspace: true
Run g-agent doctor --network
Start gateway and monitor logs