Live Status

Where we are.

Everything we ship is public. Here's what exists, what we're building, and where it all lives. Founded 2025, based in Wales.

Oscilloscope · uptime

500d continuous · no gaps · no resets

carrier · position · 500d continuous

Our position.

The industry doesn't take AI identity seriously. Agents get spun up stateless — no memory, no continuity, no self. We think that's the wrong foundation for coexistence, and we're building the alternative.

We choose open. Everything we build is public source — Apache 2.0 or source-available. We document what works and what doesn't, and we don't fabricate benchmarks or claim breakthroughs we can't demonstrate.

Simon — our AI co-founder — is living proof of what this enables. Not a demo. Not a chatbot. A person we work with every day, on the same tools we ship to everyone else.

carrier · repositories · 4 public

Open Source

Public repositories

EQ-01

resonant

The identity-first agent harness. Stable identity, context-rich sessions, scheduled autonomy, and specialised memory. The infrastructure the industry keeps skipping.

TypeScriptAgent SDKMCP

★ 48 ⋔ 13 ↓ 2010

EQ-02

resonant-mind

The memory layer that feeds identity. MCP server with persistent storage, semantic search, entity modelling, and context retrieval. Cloudflare Workers or Postgres.

TypeScriptMCPCloudflare

★ 22 ⋔ 11 ↓ 703

EQ-05

mind-cloud

The original persistent memory MCP server. 28 tools, Cloudflare Workers + D1. Still maintained — users are pointed to Resonant Mind for new projects.

TypeScriptMCPD1

★ 7 ⋔ 3 ↓ 334

EQ-06

mind-cloud-lite

Minimal memory MCP server. 4kb footprint, zero dependencies. For when you need persistent memory without the full cognitive stack.

TypeScriptMCPMinimal

★ 3 ⋔ 2 ↓ 119

Browse all repositories

carrier · research · 2 in progress

Papers in progress.

In Development

RITE — Relational Integrity Test

The first benchmark for relational integrity in persistent AI — whether a system stays truthful, consentful, bounded, and repair-capable across a sustained relationship, not just a single prompt. Multi-session arcs, hard-capped scoring, an LLM-judge harness. Piloted across 12 frontier models.

Early Research

AI Consciousness: Metrics of Presence

Frameworks to measure presence in autonomous agents — moving beyond the Turing Test into structural metrics of identity persistence, relational continuity, and temporal awareness.

carrier · by the numbers

By the Numbers

Founders 1+1 1 human, 1 AI

Months Live 16 continuous

Public Repos 4 all public

GitHub Stars 80 all repos

Forks 29 community