Skip to content

Controls how the agent executes generated shell commands:

  • Request Review: The agent will never auto-execute terminal commands (except those explicitly added to your configurable Allow list).
  • Proceed in Sandbox: Commands run automatically inside the Terminal Sandbox; commands that need to run outside it still require review.
  • Always Proceed: The agent will execute commands automatically without prompting (except those explicitly added to your configurable Deny list).

Allows the agent to view and edit files outside of the active project folders.

  • By default, the agent only has access to the folders inside your Project and the application’s local app data directory ~/.gemini/antigravity/ (which contains Artifacts, Knowledge Items, etc.).
  • Enforcing this boundary protects your local sensitive data. Enable non-workspace access with caution.

Restricts agent terminal commands to an isolated sandbox using native operating system isolation (Linux namespaces and macOS sandbox-exec).

  • When enabled, shell commands execute inside the sandbox without access to sensitive system paths or unauthorized networks.
  • Configurable globally in application preferences or overridden per project.
  • Learn more in the Terminal Sandbox guide.