org-os — the organizational operating system

org-os is the operating system this cooperative actually runs on — and a project in its own right. The organization's entire working state lives in plain, versioned files: registries for members, projects, finances, tasks, meetings and relationships, published as machine-readable schemas under the EIP-4824 DAO standard. Nothing about how we operate is locked inside a platform; the org's memory is a repository the co-op owns.

Agents operate inside it, not on top of it. Meeting recordings become structured notes, minted decisions, and tracked tasks through a processing pipeline with named sources for every claim. A Telegram agent handles intake and routing from the team's daily channels into the same registries. Knowledge work — from scattered markdown to typed, reviewable objects — runs through the same discipline: local files first, human review gates, an auditable trail.

Why build this instead of adopting an agent platform? Because the current wave of tools hosts your operations on their infrastructure — org-os inverts that. The substrate is git, YAML and Markdown: readable by any tool, portable to any host, forkable by any sister organization. Three federated instances already run on it across our network, and the same structure is what lets agents work reliably: they read and write the same files humans do, under the same review. That inversion — your organization as a commons you own, with agents as accountable operators inside it — is the difference, and it is the operating layer beneath every service we offer.