By Vero SecurityBuilt and funded by Vero Security, the AI Security Engineering platform.Read the charter

Disabling the unsandboxed retry escape hatch

When a command fails in the sandbox, the agent may retry it with dangerouslyDisableSandbox. Set sandbox.allowUnsandboxedCommands to false to ignore the parameter, or add an ask rule so every retry prompts. In managed settings this makes the sandbox admin required, so a repository cannot loosen it.

Read in this order

  1. OfficialThe escape hatch (Claude Code sandboxing docs)Shows who approves a retry in each mode and the strict sandbox settings.
  2. OfficialSettings reference (Claude Code docs)Defines sandbox.allowUnsandboxedCommands and related sandbox settings.

Save your progress

We email you a sign-in link. No password.

By signing in you agree to the terms. The privacy policy explains how we use your details.

Join the AI security engineering community

Free. Roadmap updates, chapters near you, and first word when credentials open.

We’ll email you community updates. Unsubscribe anytime. By sending this you agree to the terms. The privacy policy explains how we use your details.