Welcome to celestia-island

celestia-island is a suite of projects for industrial AI control: multi-agent collaboration, remote operations, and safety-critical automation. This site is its why — the philosophy, the ecosystem map, and the entry point. The how lives in per-project documentation sites linked from here.

Answering three questions

QuestionWhereWhat you will find
Why does this exist?PhilosophyThe problem we are solving, the closed loop, the safety doctrine, and the long-term horizon
What is inside?EcosystemEvery project, its role in the loop, and where its documentation lives
How do I start?Get StartedThe 30-minute path from account to a working chat agent and industrial control

The one-paragraph summary

celestia-island builds the closed loop from discovery to verification for AI-driven industrial control: discover → install → authenticate → deploy a model → chat and run agents → control industrial equipment → verify everything. The loop is assembled from small, strictly layered pieces: authentication primitives (kirino), platform facilities (plana), UI components (hikari), and services that only implement business logic (arona, shittim-chest, entelecheia, evernight). Nothing is ever implemented twice: generic capability is built upstream once, and consumed by every service downstream.

The reason for all of it is a simple observation: on the Moon a round-trip takes 2.6 seconds, on Mars 6 to 44 minutes. Robots out there cannot wait for a human on Earth — they must be locally autonomous. The decision layer, world model, and safety gates we are building for industrial control today are the same shape that autonomy will need tomorrow.

Where everything lives

Use the language switcher (bottom-right) to read this site in another language. Content is authored in English; translations follow the same structure.