Setup and Modes

Prerequisites

Before using the dev container, make sure you have:

• Docker Desktop or a compatible container runtime
• VS Code with the Dev Containers extension

Quick Start

1. Export the required host environment variables.
2. Create ~/.config/opensana on your host if it does not already exist.
3. Open the repo in VS Code.
4. Run Dev Containers: Reopen in Container.
5. Open a terminal inside the container and start the AI CLI you want to use.

Required Host Environment

For the current Bedrock-backed Claude Code flow:

export AWS_BEARER_TOKEN_BEDROCK=<your-aws-bearer-token>

Optional variables include:

• AWS_REGION
• CLAUDE_CODE_USE_BEDROCK
• CLAUDE_CODE_MAX_OUTPUT_TOKENS
• MAX_THINKING_TOKENS
• DEVCONTAINER_NETWORK_MODE

Network Modes

The dev container supports two modes:

• strict: enables the repo firewall and blocks outbound traffic by default
• open: skips the firewall and allows normal outbound traffic

Set the mode on your host before opening or rebuilding the container:

export DEVCONTAINER_NETWORK_MODE=strict
# or
export DEVCONTAINER_NETWORK_MODE=open

Why The Container Helps

The container isolates:

• Linux-native dependencies in /workspace/node_modules
• AI CLI configuration and session state
• optional network access rules

That makes it a better fit for controlled AI execution than mixing everything into the host machine.