Fifteen machines between a garage and an office, two NAS boxes, a custom water loop, and a 10-gigabit llama.cpp cluster running a 212B. Roughly two million tokens of live context across three brain nodes. Local inference, persistent memory, no rented GPUs. This is the build log.
Downloads are Hugging Face rolling 30-day totals across published Robinson Labs models, a public counter we do not control. A merge is agent-authored when the PR body carries the Generated-with-Claude-Code trailer left by the pr-architect flow. Counts cover Robinson Labs own-org repositories only, and every merge is human-reviewed before it lands, so an agent opens the PR but a person makes the call. Numbers refresh at deploy time.
★ per-node star assignments are illustrative, firm names land at each rebuild (DEC #555). Constellation names are locked.
Robinson Labs is the homelab of Stephen Robinson, a ServiceNow architect in Charlotte, NC who got tired of renting intelligence. By day I design enterprise platforms; by night I run a local-first AI operating system on hardware I own.
The bet is simple: own your models, your memory, and your tools. Local inference, persistent memory across sessions, agentic tooling that does real work, no rented GPUs, no API key on the critical path. The interesting problems live in the wiring, not the billing.
Right now the cluster is a pile of PCs under my desk; tomorrow it moves to the garage when the last 10-gig SFP+ lands. This is where I write down what I learn building it: the dead ends, the BIOS resets, the "nothing changed but everything broke" nights.