Skip to main content
The weekend build (Steps 0–4) is the whole foundation. Everything below is optional and should be added only when a real need pulls it in, in roughly this order. Adding these early is the over-engineering this design exists to avoid.

The order

  1. Observability — once you have services worth watching. Metrics, logs, and traces so you’re never guessing what your own box is doing. Shipped here as a clearly labeled optional module: optional/observability/. Read its README before you turn it on; the point is that you don’t start with a monitoring stack.
  2. More tools — when you hit a wall a new MCP server removes. Add the one that unblocks you, not a catalog.
  3. A second agent / model — when one model’s weakness starts costing you real time. The gateway already makes this a config change.
  4. Databases — when something needs to persist structured state (a real app, a knowledge base with vectors). Run them in Docker, bound to the tailnet.
  5. A deploy path — when you’re shipping public sites. Serve them through an outbound tunnel (e.g. Cloudflare Tunnel / cloudflared) so you keep zero ingress: the box dials out, nothing dials in.
  6. GPU, rented by the job — the first time a task genuinely needs a card. Burst to a serverless GPU backend for the length of the job; don’t buy one to idle. See WHEN-NOT-TO.md for the exception (24/7 GPU-bound work).

Things people build on this

  • A coding environment any model can drive.
  • A private second brain — a knowledge base your agents read and write, on hardware you control.
  • A prototype factory with real CI and a real deploy path, no per-experiment cloud bill.
  • A research rig that borrows GPU muscle by the minute.
  • A self-hosted automation hub — watch a source, summarize, act, notify.
  • A place to actually learn infrastructure by running it.
Pick one. The house doesn’t care which. That’s the point of owning it instead of renting it furnished.

The rule

Do not front-load the ambitious parts. Do not stand up a monitoring stack before you have three services to monitor. Do not run two deploy controllers because both looked interesting. Do not buy a GPU. Complexity added before you need it isn’t a head start — it’s debt with no upside yet.