As the primary motivation for a sandbox is to run agents with permissions disabled safely, it should be clear from the documentation how to do it.
What I tried and it worked is to add the following into .openlock/config.yaml for Claude Code:
args: ["--dangerously-skip-permissions"]
As the primary motivation for a sandbox is to run agents with permissions disabled safely, it should be clear from the documentation how to do it.
What I tried and it worked is to add the following into
.openlock/config.yamlfor Claude Code: