🪞 Intelligent Proxy · Digital Twin

Digital Twin

When You're Away, Your AI Twin Holds the Fort.

Create your AI digital twin: automatically handles messages when you're away, preset departure instructions keep your twin informed of your intent, AI drafts replies in your style, and forwards to designated bots when it can't handle things independently.

Draft Lifecycle

readysendingexpired / failed
💡 All twin learning and inference happens locally by default. Your conversational style data never leaves your device.
🎭

Thinks and Speaks Like You

Based on your historical conversations and user profile, the twin's drafts closely mirror your personal wording, tone, and logical structure.

🛡️

You Always Hold Final Control

Drafts require your review and confirmation by default. The twin won't overstep by auto-sending sensitive content. All actions have complete audit logs.

📡

Smart Forwarding When Stuck

The twin knows its capability boundaries. For complex or sensitive messages it proactively calls forward_to_bot, ensuring no important message falls through the cracks.

01

Configure Twin Profile

Write a system prompt, set departure instructions, upload historical conversation samples so AI fully understands your style.

02

Activate Twin Mode

Once activated, all messages from connected platforms automatically enter the twin processing queue — no manual intervention needed.

03

Draft Generation & Review

AI generates reply drafts based on message content and your historical style. Draft status is visible in real-time.

04

Send or Forward

Drafts automatically enter the send queue. Messages beyond handling scope are forwarded to designated bots via forward_to_bot.

Core Features

A genuine extension of yourself, not just a simple auto-responder

🪞

Digital Twin Profile

Write a dedicated system prompt for your twin defining personality, permission boundaries, and knowledge scope — the more detailed, the more it sounds like you.

📝

Departure Instructions

Set departure instructions before you leave: inform the twin of current priorities, special handling rules, and escalation logic. Rich text format supported.

✍️

Style-Matched Draft Generation

AI analyzes your historical message style and auto-generates reply drafts matching your expression habits for every incoming message, maintaining consistency.

🔀

Forward-to-Bot Tool

When the twin encounters messages beyond its scope, it invokes the forward_to_bot tool to forcibly forward to a pre-configured specialized handling bot.

📊

Draft State Management

Every draft has a clear lifecycle: ready → sending → expired / failed. Manual intervention supported at any stage.

🧹

Expiry Scan & Auto-Cleanup

Scheduled scan tasks automatically mark unsent drafts that have timed out as expired, cleaning invalid queues and keeping the twin system running efficiently.

Pair with IM Integration to build a complete intelligent message ecosystem: