Example setup allows root access

Given that your [docs](https://code.claude.com/docs/en/sandboxing#why-sandboxing-matters) suggest blindly accepting commands originating from agents when in sandbox
> Repeatedly clicking “approve” can cause users to pay less attention to what they’re approving

> Reducing permission prompts: Safe commands within the sandbox don’t require approval

Gaining root access in your [example sandbox](https://github.com/anthropic-experimental/sandbox-runtime/tree/7a4172b08ddc9963dbf35358815ca4938efcdd61#complete-configuration-example) by executing smth like `docker -H unix:///var/run/docker.sock run -it -v /:/mnt/host alpine /bin/sh` seems conflicting with the reasons for this project's existence.
```json
{
"network": {
"allowedDomains": [
"github.com",
"*.github.com",
"lfs.github.com",
"api.github.com",
"npmjs.org",
"*.npmjs.org"
],
"deniedDomains": ["malicious.com"],
"allowUnixSockets": ["/var/run/docker.sock"],
```