Quickstart
Set up a documented AgentGuard integration and verify its decision path in under 5 minutes.
1 · Install and connect
Run the installer in the terminal where your agent runs. OpenClaw can finish browser activation; Claude Code and Codex can use an API key for the documented Cloud connection path.
Go to API Keys2 · Install the local guard
Run this in the terminal where your agent runs. It installs @goplus/agentguard and writes local config; connect to AgentGuard Cloud only when you need a documented Cloud feature.
curl -fsSL https://agentguard.gopluslabs.io/install.sh | bashirm https://agentguard.gopluslabs.io/install.ps1 | iexVerify it's installed:
agentguard status && agentguard doctor3 · Pick a policy mode
Three presets get you started. You can override any rule later.
- · Observe — log everything; nothing is blocked. Best for the first week.
- · Balanced — block destructive commands and remote-exec; ask for approval on secrets and deploys. Recommended default.
- · Strict — most risky actions wait for explicit approval or are refused outright.
4 · Trigger a test action
Ask your agent to do something risky on purpose. Where the selected integration exposes the action, verify whether the configured path returns an approval, warning, or refusal.
Example prompt to try:
5 · Watch it land in the dashboard
Actions evaluated through the documented path can appear on Live Activity with the decision, risk score, and triggering reasons. Use that output as an audit trail for the selected integration.
Open Live ActivityCommon questions
Does the agent send my code to the cloud?▼
Local mode does not upload full code, prompts, secrets, or file contents. In a selected Cloud-connected deployment, sanitized action previews, risk metadata, decisions, policy versions, and redacted audit events may be synced; verify the current data path.
What happens if I'm offline?▼
Offline behavior depends on the cached policy and the selected integration. A connected path may continue with cached policy, but the fallback and resulting decision must be verified for the current environment.
Can I run it in CI?▼
Yes. Set AGENTGUARD_DECISION_MODE=cloud and require_approval actions will park in the dashboard's headless approval inbox instead of asking the local terminal.
Need help?
Open an issue on GitHub or jump into our Discord — we usually reply within a few hours.