The Intent Layer

Open-source infrastructure for AI partnerships that persist.

The Problem

Your AI forgets you every conversation. Your preferences, your working style, the decisions you made yesterday. Every session starts cold. You repeat yourself. The AI guesses.

The Solution

The Intent Layer gives any AI client a persistent identity layer: who the human is, how they work, what's happening, and how the AI should behave. One system, any model, every conversation starts warm.

Products

AI Command Center

Template repo for your identity rules, personalities, and status tracking.

View on GitHub →

MCP Server

Runtime API that serves your identity data to any MCP-compatible AI client.

View on GitHub →

This Website

OAuth provider for MCP connectors. Guided onboarding and configuration dashboard.

Request early beta access →

How It Works

Three layers define your AI partnership:

1
Identity Rules Who you are, how you work, what you expect from AI.
2
Personalities Behavioral modes the AI switches between based on context.
3
Status What's happening now -- active work, deadlines, blockers.

Architecture

Your AI Command Center repo stores your identity data in GitHub. The MCP server reads it and serves it to any MCP-compatible client (Claude, Copilot Studio, or any tool that speaks the protocol). This website handles OAuth authentication so Claude can connect securely, and provides a dashboard for managing your configuration.

A
GitHub Repo Your identity rules, personalities, and status files live here.
B
MCP Server Reads your repo and serves identity data via the MCP protocol.
C
This Website OAuth provider + dashboard. Connects Claude to your MCP server.

Related

AI Trust Commons

Open standards for trustworthy AI partnerships.

aitrustcommons.org →

HIP Charter

Human-Intent Preservation charter for AI systems.

hipcharter.com →

Founded by Nikhil Singhal