Public Intelligence

An OpenAI-compatible control plane for hardware you already own.

Point it at machines you or people you trust control — a workstation with a GPU, a lab's spare box, three laptops in a research group — and it gives you one authenticated /v1/chat/completions endpoint that routes across all of them, including the ones behind NAT, with no port forwarding and no VPN.

It is not a marketplace and there is no network to join. You run both halves. See docs/decisions/D8-the-wedge.md for why that is the product, and D2 for the arithmetic that ruled out the alternative.

This is pre-alpha. That sentence is doing real work — read "What is actually true" before relying on anything here.

What is actually true, as of 2026-08-07

Verified, not aspirational. STATUS.md is generated from real test runs; this section summarises it in prose.

Works: a node registers with the Scheduler, heartbeats over an authenticated Zenoh mesh, advertises the models Ollama actually has, and survives the Scheduler being unreachable. The Scheduler matchmakes, dispatches over the mesh (which is how a node behind NAT is reachable at all), and exposes an OpenAI-shaped gateway behind RS256 JWT auth. State can persist across restarts. CI runs the gate on Linux, macOS and Windows across Python 3.11–3.14; STATUS.md records the last measured result and the commit it covers.

Does not work, and is not claimed to:

• No node on a genuinely separate machine has ever served a request. The mesh transport is built and tested against a real Zenoh router, but in one process on loopback (ROADMAP.md, 1.5). This is the single most load-bearing unproven claim in the project — see docs/PREMISES.md, P2.
• Split inference, layer sharding, FP8 compression, speculative decoding and KV-cache checkpointing are not implemented. They are cut from v1. The gateway answers 501 if you ask for split inference. It used to answer 200 with invented text; that was removed on 2026-08-07.
• Credits are an accounting unit, not a currency. They are not redeemable and there is no payout path — not "not yet", but a decision (D2).
• There is no content filtering, no meaningful rate limiting, and no backups, and nothing verifies that a node ran the model it claims. docs/OPERATING.md lists what you take on by running this.

What this project has decided, and what it hasn't

The eight product questions this code had been assuming answers to are answered in docs/decisions/. The short version: invite-only trusted hosts, no marketplace, no operated network, one coordinator, and a self-hosted pitch.

One question is still open and cannot be closed from inside: D7 asks for a second pair of eyes. Every judgement here has been made by one party, and the verification process — strong as it is — is structurally incapable of catching a wrong premise. The assumptions are laid out in docs/PREMISES.md specifically so they can be attacked. If you disagree with one, that is the contribution.

Repository layout

Path	What it is
packages/scheduler/	FastAPI control plane: registry, matchmaking, OpenAI gateway, Zenoh router
packages/node/	FastAPI host agent: local control API, telemetry, Ollama-backed inference
packages/website/	Next.js dashboard and playground
specs/	One document per change: what it does, why, and what is out of scope
docs/decisions/	The product decisions, and what each one costs
docs/historical/	Superseded design documents. They describe intentions as if built.

Running it

One venv for everything:

python3 -m venv .venv
.venv/bin/python -m pip install -e packages/shared \
  -e "packages/node[dev]" -e "packages/scheduler[dev]"
./scripts/install-hooks.sh
./scripts/verify.sh          # the single gate: lint, types, tests, security, installer

scripts/verify.sh is the only definition of "does this pass"; CI runs that file and nothing else. See CONTRIBUTING.md for how to work here and VERIFY.md for what a completion claim has to be backed by.

Licence

Apache-2.0. Patent grant included; warranty disclaimed. Security reports go through SECURITY.md, not the issue tracker.