NamerynAI Agent Rules Builder
Type what your agent does, one task per line, and get one narrow rule per action with the setting to pick in dots, Grok Bot and Muse. Money, outgoing messages and deletions always ask first. Keep a private record of which agent can read or change which app, and a checklist for taking it back.
What each agent can touch
| Agent | App | Access | Granted | Revoked |
|---|
Kept in this browser only. Tick Revoked once you have removed the access in the agent's own settings.
FAQ
How do I set rules for my AI agent?
Write one rule per action, starting with the verb: Pay my phone bill, Send the weekly report, Summarize unread email. Then pick how careful the agent must be with each one. This page suggests the setting by kind of action and refuses to let money, outgoing messages or deletions run without asking.
What rules can dots follow?
For each action you describe, OpenAI lets you pick one of four behaviors: take action without asking, take action if pre-approved, ask before taking action, or hand off to you. Pre-approved means you asked for the action in your prompt.
How do Grok Bot's approvals work?
Grok Bot asks before running on your own computer by default, and lets you write personal rules of two kinds, Ask first and Allow automatically; when both match, Ask first wins. xAI suggests narrow rules around a known action and scope.
Can Muse act without asking me?
Meta says Muse's Sentinel decides, from the policy you set, whether an action is allowed, denied or needs your approval, and that you can grant a permission once, for a session, for a task, for a set time, or for good.
Where is my access record kept?
In this browser only. Export it as a CSV to keep a copy; nothing is sent anywhere.