Every time you open a new AI chat, you start from zero. mytwin.space fixes that. Build a structured identity — voice, values, goals, constraints, public & private memory — and connect it to any MCP-enabled AI. Your context loads automatically, with permissions enforced. Includes an identity editor, simulation mode (“how would I respond?”), version history, and an MCP server with scoped API tokens. Early beta. Core loop is live.

AppSignal — Full-stack monitoring for errors, metrics, and logs

Full-stack monitoring for errors, metrics, and logs

Hey PH 👋 I built mytwin.space because I was tired of explaining myself to AI from scratch every session. The insight: your identity is structured data. It has fields (voice, values, constraints, goals), it should be versioned, and it should be permissioned. So I built it that way. The MCP integration is the part I'm most excited about. Once you set up your twin token in Claude Desktop, your context loads automatically at the start of every conversation — no copy-pasting, no prompt engineering. The AI just knows who you're working with. What I'd love your help validating: 1. Is the "twin" mental model intuitive on first contact? 2. Does the MCP angle resonate, or is it too technical for most users? 3. What would make you actually use this daily? Happy to answer anything. The beta is gated but I'm approving everyone who signs up today within the hour.

AppSignal — Full-stack monitoring for errors, metrics, and logs

Full-stack monitoring for errors, metrics, and logs

My Twin Space

Build your AI identity. Connect it to any model.

Build your AI identity. Connect it to any model.