import type { Metadata } from "next"; export const metadata: Metadata = { title: "Infrastructure — Powered by MacLustr", alternates: { canonical: "/infrastructure" }, }; export default function InfrastructurePage() { return (

Powered by MacLustr

Research computed on MacLustr

AutoScholar is the research startup. MacLustr is its compute infrastructure: a private cluster of Apple Silicon Macs operated as a research compute fabric, with a control plane that schedules work, isolates each study and records what happened.

{[ [ "Apple Silicon nodes", "Research jobs run on Apple Silicon Mac nodes with unified memory, well suited to numerical work, local model inference and LaTeX builds.", ], [ "Isolated per-run workspaces", "Each run executes under a dedicated low-privilege service identity in a fresh workspace with OS-enforced sandboxing, restricted egress, and time, disk, memory and budget limits.", ], [ "Durable scheduling", "A durable queue with leases, heartbeats and fencing tokens means a run survives page reloads, network blips and worker restarts. Expired workers can never publish stale results.", ], [ "Evidence first", "Every artifact is hashed and versioned on write. Publication is a separate service the research agent cannot call.", ], ].map(([h, p]) => (

{h}

{p}

))}

What we publish about the infrastructure

We credit MacLustr on every paper and describe its design honestly. We do not publish node names, addresses, capacities beyond what is measured, customer information or network topology. Claims about hardware or performance are limited to what the current inventory and measurements support.

Model execution

Research runs use the Claude Agent SDK with API-key authentication. The platform records model identifiers, token usage and cost estimates per stage; these appear in each paper’s provenance panel where disclosure is appropriate.

); }