jakman85andClaude Opus 5 3ade1d440f fix(conformance): status.py read its own detection vocabulary as a TODO marker
`has_todo = "TODO" in text` — a bare whole-file substring. The new L0 check
`placeholder-guard-coupling/check.sh` ENUMERATES placeholder idioms as data
(`SEED = ["TBD","TODO","FIXME",...]`) because detecting them is its entire job, so
the harness classified a 32-assertion executable check as a TODO scaffold.

Consequence: R-STATE-03 — SAFETY-CRITICAL — reported `exec=1 scaffold=1` and
`not-derivable`, because status.py requires all of a requirement's tagged tests to
be proven. A safety requirement mis-reported by a substring match, plus a standing
--strict-coverage conflict that a previous engineer worked AROUND by rewording
prose instead of fixing the detector.

Replaced with two REDUNDANT structural signals, derived from the corpus (all 43
tests read), not invented:
  1. `# STATUS: scaffold` with the state as the FIRST token after `STATUS:`. Exactly
     3 tests declare it — onboarding-completeness, alerting, gitops-drift — and they
     are exactly the three real scaffolds. 19 others declare EXECUTABLE/RUNNABLE, so
     a loose `.*scaffold` would have matched prose.
  2. A TODO/FIXME marker CARRYING A MESSAGE in the body, where the body starts after
     the header comment block. A marker that is the ENTIRE content of a quoted
     literal is data, not a note, and is blanked before the scan. Every genuine
     scaffold marker in the corpus carries a message, so mid-body comments still count.

The line-322 intent is preserved exactly: each file is classified on its own
structure, either signal in either file sets scaffold + annotation_conflict, the
`executable: "true"` annotation is consulted only AFTER the body verdict, and
ambiguity still only downgrades. Vocabulary stays TODO/FIXME only — TBD/PLACEHOLDER
are excluded because two checks use them in real executable code.

NON-VACUITY, and the redundancy is the point: deleting the `# STATUS: scaffold` line
from all three real scaffolds still catches all three via the body rule. The decisive
perturbation is the reverse — the exact file that caused the bug, degraded to a
TODO-only body while KEEPING its `STATUS: EXECUTABLE` header, its idiom-naming header
prose and its `executable: "true"` annotation, is caught as a scaffold WITH the
annotation conflict raised. The annotation still cannot outrank the body.

TAGGING: the agent RECOMMENDED KEEPING `R-STATE-03` on placeholder-guard-coupling and
reversed my lean, on a ground I had missed and which is load-bearing:
ci/touched-requirements.sh maps `bootstrap/*` -> R-STATE-03, and `run.sh
--requirements` resolves R-IDs to test dirs THROUGH THESE LABELS. Removing the tag
would reopen the exact merge-gate hole a984442 closed. It also measured the
counterfactual — with the tag removed R-STATE-03 becomes `stale-proof`, not `proven`
— so the tag is not what holds PROVEN at zero.

R-STATE-03 before -> after: `exec=1 scaffold=1` / not-derivable + strict conflict, to
`exec=2 scaffold=0` / not-derivable with an ACCURATE reason (no ledger row for the L0
check; dr-restore is stale-proof).

LEDGER MECHANISM CONFIRMED CORRECT, not worked around. R-STATE-02's node-loss proof
(sha 57d4df8c) went stale because a984442 touched its evidence set — attributed
precisely to the single path `ci/touched-requirements.sh`. Fail-closed and by design.
Re-proving needs a real L3 run recorded from a clean tree at HEAD-or-later: a
live-cluster action.

OBSERVATION FOR LATER (not changed): scan_implementation counts
ci/touched-requirements.sh as implementation evidence for EVERY R-ID it names, so any
edit to that dispatch table invalidates every proof it mentions. Both current
stale/not-derivable verdicts trace to that one file. Fail-closed is the right
direction; whether a CI path-map should be implementation evidence is a separate call.

VERIFIED BY ME: `status.py --selftest` PASS (8 new assertions, both directions);
`--status --strict-coverage` exit 1 -> 0; `--layer l0` 11/11; the three
`STATUS: scaffold` files are exactly the three real scaffolds; kyverno L0 29/29 and
the full L0 stage green IN THE CONTAINER (the agent saw the stage skip on the host,
where the pinned kyverno CLI is absent). Full CI gate PASS. Counts unchanged:
coverage 48/51, scaffold-only 4, PROVEN 0 (stale 1).

Refs R-TEST-01, R-TEST-04, R-TEST-05, R-STATE-03.

Co-Authored-By: Claude Opus 5 <[email protected]>
2026-09-06 20:24:22 -05:00

home-network-k3s-impl

Implementation of the home k3s + Longhorn platform. This repo holds the reconciled artifacts (Flux config, Kustomize overlays, KubeVirt/Helm manifests, ServiceDefinitions, Kyverno policies, Chainsaw conformance tests, CI). It is driven by — and subordinate to — the read-only specification at ../home-network-k3s.

Rules (from the spec — non-negotiable)

  • The spec is read-only. Never modify anything under ../home-network-k3s. Implement against it; escalate spec problems, don't edit them (spec AGENTS.md §11, R-TEST-06).
  • Done = a green, reproducible, requirement-tagged conformance test — never "merged," never a manual demo (spec docs/14, agents/definition-of-done.md).
  • No work on main. Every change is a feature branch merged into main (spec CONTRIBUTING.md).
  • No plaintext secrets — SOPS-encrypted only (spec docs/07).
  • Phase discipline — only build what the active roadmap phase authorizes (spec roadmap/phases.md).

Layout

infra/          node provisioning: OpenTofu (Proxmox VE) + cloud-init + Ansible, and the
                pinned dev container that carries the whole toolchain
bootstrap/      one-time cluster bootstrap: Flux install, SOPS/age key, host Gitea
platform/       platform CONTROLLERS (Flux HelmReleases / operator bases): kyverno, cilium,
                longhorn, external-snapshotter, cert-manager, externaldns, gateway, coredns,
                vpn-egress, monitoring, cnpg, keycloak, oauth2-proxy, kubevirt
platform-config/ platform CONFIGS — the custom resources of the CRDs installed above
                (ClusterIssuers, Gateway, Kyverno policies, KubeVirt/CDI CRs, Keycloak CRs).
                Split out because CRs and the CRDs they use cannot apply in one pass.
apps/           application tier, reconciled as dependency-ordered WAVES:
                  wave-0/ wave-1/ wave-2/   rendered container apps (see DESIGN-waves.md)
                  vms/                      KubeVirt VirtualMachines (non-gating tier)
catalog/        ServiceDefinition entries (the onboarding surface) + render.py
schemas/        ServiceDefinition JSON Schema + renderer contract
clusters/       Flux entrypoints + dev/ and prod/ overlays
conformance/    Chainsaw behavior/conformance suite + harness + coverage report
ci/             the gate as scripts (lint · schema · Kyverno L0 · L1 render · decision)
.gitea/         Gitea Actions CI running that gate on an in-cluster runner

Status

Roadmap phases 0–6 are delivered and deployed on the dev cluster; Phase 7 (prod) is not open — clusters/prod is a design-only overlay.

Phase Delivered
0 Foundations ServiceDefinition schema + renderer, Kyverno L0 gate, conformance harness, CI
1 Cluster bring-up 3-node k3s HA on Proxmox VE (192.168.1.50/.51/.52), Cilium, Flux, SOPS
2 Platform core Cilium · Longhorn · cert-manager (LE-prod wildcard) · ExternalDNS · Gateway API · split-horizon CoreDNS
3 Egress / VPN ADR-0015 node-level OpenVPN split-tunnel + Cilium egress-gateway steering + host kill-switch
4 Observability kube-prometheus-stack · Loki · Alloy · Alertmanager. Velero still blocked on OQ-BKP; image-automation authored but disabled
5 IaaS KubeVirt + CDI on real nested KVM; live-migratable RWX VM storage. The UniFi VM runs, but UniFi OS Server is not yet installed in the guest
6 Identity + CI Keycloak (Operator + CNPG) · oauth2-proxy forward-auth · Grafana OIDC · kubectl OIDC SSO with group→RBAC · in-cluster Gitea Actions merge gate

Conformance: 36 / 51 requirements covered (70.6%), 15 gaps, 1 safety-critical gap (R-STATE-02, Longhorn node-loss). Regenerate with ./conformance/run.sh --coverage-only; the full report is conformance/_artifacts/coverage/coverage.md.

The running narrative — decisions, root causes, incidents, follow-ups — is DECISIONS-LOG.md. It, not this table, is the authoritative record.

Working on this repo

Everything the gate needs is baked into the pinned dev container — do not install the toolchain on the host:

infra/devcontainer/dev.sh build            # once
infra/devcontainer/dev.sh run make -C ci gate   # the same gate CI runs
infra/devcontainer/dev.sh                  # interactive shell (sops, flux, kubectl, tofu, ansible)

kubectl/flux/helm in there talk to the dev cluster by default. Prefix CLUSTER=prod to target prod instead:

infra/devcontainer/dev.sh run kubectl get nodes                # dev (default)
CLUSTER=prod infra/devcontainer/dev.sh run kubectl get nodes   # prod

Both clusters are always present in the container's KUBECONFIG as the contexts k3s-dev and k3s-prod; CLUSTER picks which one is current, and each run prints the cluster it is targeting.

The container mounts the repo at /work, your age keyring, and the sibling spec repo read-only so the spec-immutability check can actually run. See infra/devcontainer/README.md.

A green run prints MERGE GATE: PASS. The L2/L3 conformance layers stay SKIPPED unless CLUSTER_AVAILABLE=true — the static gate never pretends to have run them.

S
Description
No description provided
Readme
7.2 MiB
0 Stars 2 Watchers 0 Forks
Languages
Shell 61.4%
Python 32.3%
Jinja 2.9%
Dockerfile 2.2%
HCL 1.1%
Other 0.1%