YOLO Mode
Flip one switch to launch supported agents with their auto-approve flags enabled — ideal for unattended, fully autonomous runs.
What YOLO Mode does
Terminal agents normally pause to ask for permission before editing files, running commands, or calling tools. YOLO Mode injects the correct auto-approve flag for each agent, so it proceeds without prompts. That makes it perfect for background automation, long-running refactors, or CI-style workflows — anywhere you'd otherwise be babysitting the terminal.
CLI Launcher picks the right flag per agent, so you never hand-edit a launch command. Toggle YOLO Mode on for any supported agent and launch it as usual.
Example: Claude Code
With YOLO Mode on, CLI Launcher launches Claude Code with its skip-permissions flag:
claude --dangerously-skip-permissionsPer-agent flags
Different agents expose different flags. CLI Launcher maps them automatically; the most common:
| Agent | Auto-approve flag (injected) |
|---|---|
| Claude Code | --dangerously-skip-permissions |
| OpenCode | --dangerously-skip-permissions |
| Codex CLI | --dangerously-bypass-approvals-and-sandbox |
Flags for other agents follow the same pattern and are applied by the panel automatically.
Because YOLO Mode auto-approves every action — including file edits and command execution — run it only against code you trust, ideally in an isolated or disposable working directory. It is not a substitute for reviewing changes.
Turning it off
Toggle YOLO Mode off and the agent launches normally, prompting for each action. You can switch modes between launches at any time; it doesn't change your installed agents or their configuration.