[p0][Master Epic] Bluefin contributor homelab experience

[castrojo]

Bluespeed's first and primary use case: make it trivially easy for any Bluefin contributor or user to run local hardware testing, send reports back, and have Hive triage the results.

The loop:

1. Contributor runs ujust report — collects device info, stores in their own gist (user-controlled, never leaves their hands without consent)
2. Hive triages incoming reports as issues
3. Contributors can locally confirm a PR fix on their own hardware
4. Anyone with a laptop can participate — no homelab required

Why this is p0: This accelerates Bluefin contributions immensely. It closes the gap between "something is broken on my machine" and "here is a reproducible report with a confirmed fix." The entire Bluespeed platform exists to serve this loop.

The platform: Flatcar + k3s + KubeStellar Console + Argo Workflows — all deployed via just setup on a single control box. The dashboard (Bluespeed) shows fleet health, cluster state, and contributor activity. Clients are laptops running bootc images — not k3s nodes.

CNCF context: OTel Collector (incubating) ships telemetry. Prometheus (graduated) stores metrics. bootc (sandbox) manages OS updates. Argo Workflows (graduated) runs build pipelines. k3s is the distribution.

Key constraint: Laptops are clients that consume bootc images — they are never k3s nodes. Enrollment = OTel agent installation only.

[castrojo]

Ok I am working on this, some thoughts. Bluespeed is the core, flatcar with kubestellar ready ootb via another iso, which uses knuckles but is more automated for end users who just want to pick one box as their control center. Which we'll nickname ghost.

The first use case is bluefin contributor/tester. This is our first persona - we will use it to make the product better. This will cover local-first testing for everyone, so that bluefin users can run ujust commands and send reports back with a "ujust report" like feature, all their data is user controlled via their gist. Then hive triages the issues as they come in, and then contributors can test the PR and locally confirm, etc. We want hive to run the "hive" of bluefin computers.

This also means that anyone with a laptop can do this - hence we should prioritize this, it could accelerate bluefin contributions immensely. Make this p0, with the platform being built in order to achieve this goal.