Installation

Prerequisites

Before installing SCC, ensure you have:

Requirement	Minimum Version	Check Command
Python	3.10+	python --version
uv (recommended)	Latest	uv --version
Container runtime	Docker Engine 20.10+	docker --version
Git	2.30+	git --version

Container Runtime

SCC requires a Docker-compatible container runtime. Any of these work:

Runtime	Platform	Notes
Docker Engine	Linux, macOS, Windows	Standard Docker daemon
OrbStack	macOS	Fast, lightweight Docker alternative
Colima	macOS, Linux	Open-source Docker runtime
Docker Desktop	macOS, Windows, Linux	Full GUI experience
Podman	Linux, macOS	🔄 Planned — not yet fully validated

SCC auto-detects the available runtime. Docker Desktop is not required.

Caution

A container runtime must be installed and running before using SCC. Ensure the daemon is active (docker info should succeed).

Note

Installing a container runtime gives SCC a sandbox boundary, but it does not automatically restrict outbound traffic from agent sessions. Network control is configured separately through SCC’s network_policy settings.

Installing uv

If you don’t have uv installed yet, get it from astral.sh/uv:

macOS/Linux

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows

powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

pip

pip install uv

Install SCC CLI

uv (recommended)

uv tool install scc-cli

pipx

# pipx installs in isolated environment
pipx install scc-cli

pip

pip install scc-cli

Verify Installation

1. Check SCC is installed

   scc --version

2. Run diagnostics

   scc doctor

   This checks:

   • Container runtime is running and accessible
   • Runtime version meets requirements
   • Git is configured
   • Config directories exist
   • Provider image availability and auth readiness

3. Check a specific provider

   scc doctor --provider codex

   If a provider image is missing, doctor shows the exact build command.

4. Review any warnings

   If scc doctor shows issues, address them before proceeding.

Platform-Specific Notes

macOS

• Install Docker Engine (via OrbStack or Colima) or Docker Desktop
• OrbStack is recommended for speed and low resource usage
• Ensure the daemon is running (docker info)
• No additional configuration needed

Linux

• Docker Engine is the standard choice
• Ensure your user is in the docker group:

  sudo usermod -aG docker $USER
  # Log out and back in for changes to take effect

Windows (native)

• Windows Terminal, PowerShell, Git Bash, and VS Code terminal are supported
• Install Docker Desktop (WSL2 backend recommended) and keep it running
• Use uv for the most reliable install and upgrades

Windows (WSL2)

• Run SCC inside WSL2 for a Linux-like environment
• Install Docker Desktop with WSL2 backend and enable WSL integration
• File operations are fastest inside the WSL filesystem (~/projects)

Tip

For best performance, keep your projects inside the WSL filesystem instead of /mnt/c/.

Container Images

SCC builds provider-specific container images on first use:

Image	Purpose
scc-agent-claude	Claude Code agent with safety wrappers
scc-agent-codex	Codex agent with safety wrappers
scc-base	Shared base image with safety evaluator
scc-egress-proxy	Squid proxy for web-egress-enforced policy

The first scc start with a provider triggers an automatic image build (takes a few minutes). Subsequent launches use the cached image.

Manual build (if needed):

docker build -t scc-agent-claude:latest images/scc-agent-claude/
docker build -t scc-agent-codex:latest images/scc-agent-codex/

Updating SCC

uv (recommended)

uv tool upgrade scc-cli

pipx

pipx upgrade scc-cli

pip

pip install --upgrade scc-cli

Or use the built-in update command:

scc update

Next Steps

Now that SCC is installed, configure it and run your first session →