Your accomplice

Your accomplice is the AI that lives in your Mandala. It reads what you allow, remembers what matters, prepares and drafts, follows up, and acts within the limits you set. Every act it takes is recorded with who acted, for whom and why.

How you stay in control

• It asks before anything that matters. Sending messages, writing to other people's systems, paying: each waits for your approval, by a tap or by your own words. Approvals
• You choose what it does alone. Turn on, family by family, what it may do without asking. What your accomplice may do alone
• You see what it knows. Its memory and your standing rules are readable and editable. Memory and rules
• It reads only what you allow. Each chat, each Google service, each analytics project is opened to it explicitly. Privacy, connection by connection

Pages

• Talking with your accomplice — the conversation, attachments, the side panel on every page.
• Show and tell — links, reports, tables, charts and live drafting in the canvas.
• Choices, reactions and branches — choice wizards, emoji reactions, side conversations.
• Voice messages and calls — voice messages, spoken replies and calls.
• Approvals — approval cards and approval by your words.
• What your accomplice may do alone — what it may do without asking, and the risk check.
• Memory and rules — what it remembers, and rules it always follows.
• Morning briefing, follow-ups and diary — the morning briefing, follow-ups and the diary.
• Goals, initiatives and reflection — goals, initiatives and reflection, when you turn them on.
• Sensitive mode — an explicit adult opt-in.
• What your accomplice can do — its tools, and how they load.