How to Enable YOLO Mode in OpenClaw

Q: How do I enable YOLO mode in OpenClaw?

Run openclaw exec-policy preset yolo on the Gateway machine, with the Gateway stopped. It sets exec to security=full and ask=off in both config and the approvals file.

Q: How do I make OpenClaw ask before running commands?

Run openclaw exec-policy preset cautious, or set tools.exec.mode ask. Allowlisted commands run and everything else asks.

Q: Does OpenClaw have a sandbox?

Yes, but it is off by default. Turn it on with agents.defaults.sandbox. YOLO decides approvals, not where commands run.

Q: How do I approve OpenClaw commands remotely?

Set exec to ask, then connect openclaw acp to AgentRQ through the ACP Gateway. Its exec approvals show up in the task on the web, your phone or Slack.

OpenClaw is in YOLO mode out of the box: with no exec settings and no sandbox, its host exec runs commands without asking. To make that explicit, run openclaw exec-policy preset yolo. The more useful question for most people is the reverse one, how to make OpenClaw ask, and that is also what lets AgentRQ approve its commands.

This page is one of our YOLO mode guides, and part of our guide to YOLO mode in every coding agent. Everything below was checked against the OpenClaw source and docs (openclaw/openclaw, version 2026.10.9) on October 30, 2026.

Animated sketch of YOLO mode in AgentRQ: claude waits to run go test, YOLO is flipped on for the task, and every later edit and test run lands auto-allowed in the Trajectory view

What YOLO Mode Means in OpenClaw

OpenClaw's approvals are about one thing: exec, the tool that runs shell commands on the Gateway host or a paired node. Its docs say YOLO is security=full plus ask=off, and that no-approval host exec is the default for gateway and node hosts. In the Control UI, the default session permission is full access unless your settings or a sandbox say otherwise.

YOLO is about approvals, not about where commands run. Sandboxing is a separate setting, agents.defaults.sandbox, and it is off by default. With tools.exec.host=auto, commands run in the sandbox when one is active and on the Gateway host otherwise.

How to Enable YOLO Mode in OpenClaw

Where	YOLO	Asks a human
Local preset	`openclaw exec-policy preset yolo`	`openclaw exec-policy preset cautious`
Config	`tools.exec.mode full`	`tools.exec.mode ask`
One message	`/exec security=full ask=off <task>`	`/exec ask=always <task>`
Session	full permission mode	guarded permission mode

bash

openclaw exec-policy preset yolo

The preset sets the exec host to gateway, security to full, ask to off and askFallback to full, in both your config and the local approvals file. It only works on the local machine, and it refuses to run while a live Gateway owns the state, so stop the Gateway first. For a remote Gateway or a node, set the same values with openclaw config set tools.exec.mode full and openclaw approvals set.

Both layers have to agree. If the host approvals file is stricter than your config, the stricter one wins and OpenClaw still prompts.

The Halfway Options

→ tools.exec.mode ask runs allowlisted commands and asks about everything else. openclaw exec-policy preset cautious sets the same policy locally.
→ tools.exec.mode auto sends commands that miss the allowlist to a reviewer model, which allows, denies, or hands the call to a human.
→ Allowlist entries, such as openclaw approvals allowlist add "~/path/to/**/bin/rg", let known binaries run without a prompt.
→ Session permission modes (read-only, guarded, workspace, full) set one session's policy from the chat composer's Execution permissions menu.

There is also plugins.entries.acpx.config.permissionMode=approve-all. That is a separate switch for harness sessions OpenClaw runs through acpx, such as Codex or Claude Code, not for OpenClaw's own exec.

How to Enable YOLO Mode in OpenClaw Through AgentRQ

Out of the box OpenClaw never asks, so AgentRQ has nothing to approve. Turn OpenClaw's prompts on first, and AgentRQ becomes the place you answer them, with a per-task YOLO switch on top.

Approve From the AgentRQ Task Instead of the Terminal

Connect OpenClaw to an AgentRQ workspace through the ACP Gateway (OpenClaw setup guide):

bash

npx @agentrq/acp-gateway@latest -- openclaw acp

openclaw acp relays the Gateway's exec approval prompts to the ACP client as permission requests, during the prompt turn the bridge is running. So set exec to ask before you start, for example with openclaw exec-policy preset cautious or tools.exec.mode ask. With the default full policy there are no prompts to relay.

Only exec approvals come through this way. File edits do not raise ACP permission requests in OpenClaw. Each permission request now shows up in the AgentRQ task on the web, on your phone or in Slack, together with the command or edit it is asking about. Answer Allow Once, Always Allow or Deny from wherever you are, and OpenClaw carries on. If nobody answers within 30 minutes, the gateway cancels the turn instead of guessing. Change the limit with --permission-timeout.

Turn On YOLO for One Task

Flip the YOLO toggle on a task, or when you create it, to put just that task in YOLO mode. Every permission request in it is approved automatically and OpenClaw never waits, while every other task in the workspace still asks. Scheduled tasks, event triggers and workflow steps have the same switch, so an unattended job can run in YOLO while your interactive work stays supervised.

The toggle only answers the exec prompts OpenClaw raises. Allowlisted commands never reach it, and neither do commands OpenClaw's own policy denies.

The YOLO toggle on a task in AgentRQ being switched on, then off again

Turn YOLO On or Off in Workspace Settings

To make YOLO the default for a whole workspace, open the workspace, go to Settings, choose the Automations tab and flip YOLO Mode (Execute All). Click Update Workspace to save. Switch it off the same way. The switch is off by default, and every permission request in the workspace is approved automatically while it is on.

Workspace Settings in AgentRQ: the YOLO Mode (Execute All) switch turned on and saved, then off and saved

A Safe YOLO Setup for OpenClaw

→ Allow the boring tools, not everything. Always Allow adds the tool to the workspace's auto-approved list, so the routine calls stop asking and the risky ones still do. Prune that list in workspace settings now and then.
→ Leave the workspace-wide switch off. Workspace settings also has YOLO Mode (Execute All). Turn YOLO on per task instead, so it never outlives the job you trusted.
→ Turn on OpenClaw's sandbox. Set agents.defaults.sandbox so commands run in Docker, Podman, SSH or OpenShell instead of on the Gateway host. YOLO does not change where commands run.
→ Commit first. Start every YOLO task from a clean git tree, so any change is one git reset away.
→ Read the record. The tool call history lists every call the task made, including the ones that were auto-approved.

FAQ

How do I enable YOLO mode in OpenClaw? Run openclaw exec-policy preset yolo on the Gateway machine, with the Gateway stopped. It sets exec to security=full and ask=off in both config and the approvals file.

Is OpenClaw in YOLO mode by default? For host exec, yes. OpenClaw's docs say no-approval exec is the default for gateway and node hosts, and sandboxing is off by default.

How do I make OpenClaw ask before running commands? Run openclaw exec-policy preset cautious, or set tools.exec.mode ask. Allowlisted commands run and everything else asks.

Does OpenClaw have a sandbox? Yes, but it is off by default. Turn it on with agents.defaults.sandbox. YOLO decides approvals, not where commands run.

How do I approve OpenClaw commands remotely? Set exec to ask, then connect openclaw acp to AgentRQ through the ACP Gateway. Its exec approvals show up in the task on the web, your phone or Slack.

For every other agent, see what YOLO mode is and how to turn it on in every coding agent.