Community website built with CDev (#6) — placement, stack and scope

[elmayii]

Issue #6 proposes building the community website with CDev itself: a frontend + backend pair, conditioned and developed through the loop, living as the project's canonical example. The what is settled there; this discussion is for the how. Per our change policy (CONTRIBUTING.md), new capabilities start as a Discussion — input wanted before it becomes a sprint.

Open questions:

1. Placement — examples/ directory inside this repo, or separate cdev-example-* repositories linked from here?
In-repo keeps everything discoverable; separate repos give the example its own honest git history (which is the product's proof). Leaning separate-repos + an examples/README.md index here, but unconvinced.

2. Stack — CDev is method, not stack, so the choice is free. Criteria: boring, cheap to host, easy for a reader to replay. Candidates welcome (e.g. static-friendly frontend + minimal API backend). The stack must not upstage the method.

3. Backend scope — smallest API that is still real: releases feed? field-reports index? community links? A backend with one honest endpoint beats a fake microservice.

4. Hosting — free tier, zero-maintenance preferred. The site must survive being ignored for a month (that's the point of the demo).

5. Maintenance cadence — the demo's core is the maintenance sprint: a later session continuing purely from repository state. How often do we exercise it — monthly? per release?

If you've run CDev on a real project, your field experience on any of these is exactly what this needs.