Every manual launch of launch-report/launch-upgrader bypasses the systemd EnvironmentFile (/srv/cc-ci/upgrader.env sets LOOP_TIER=go + the go models) and fell back to hardcoded zen defaults, which die instantly here (ProviderModelNotFoundError: opencode/glm-5.2 — the host holds no ZEN key and zen models vanished from the provider catalogue, 2026-09-14). This crashed the 2026-09-11 report agent at launch and again on every manual retry.
- launch-upgrader.py + launch-report.py: default TIER go, go-tier model default glm-5.3-flash (mirrors the 2026-09-08 operator choice in upgrader.env); zen stays an explicit opt-in on a host with a key.
- launch-supervisor.py: default SUPERVISOR_MODEL glm-5.3-flash.
- docstrings updated to match.
Verified with env -i: bare invocation resolves tier=go, model=opencode-go/glm-5.3-flash on all three. The 2026-09-11 weekly report is published live (report.ci.commoninternet.net/week-2026-09-11.html) after a fresh glm-5.3-flash report run.
The operator's AUTONOMIC_OPENCODE_KEY is a Go subscription key. Put in the
`opencode` (ZEN) slot it authenticates but every request fails "Insufficient
balance", because ZEN is pay-as-you-go credit; on the `opencode-go` endpoint
the same key answers fine. So the orchestrator agent moves to
opencode-go/glm-5.2, and README records that this host is a Go host and that
its opencode config must carry no inline apiKey.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqkQq3CDmFWcQ7u1LzoyRz
The 2026-09-07 weekly run's step-0 orphan sweep removed the `opencode-ui`
swarm stack (the traefik route to the opencode web UI) because it was not
on sweep-orphans.sh's keep-list. Renamed to ccci-opencode-ui alongside the
other control-plane stacks and added to KEEP_RE.
Journal: the run itself (lasuite-docs #8, n8n #7 GREEN on the new Drone,
report week-2026-09-07 published) ran entirely on the new host.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqkQq3CDmFWcQ7u1LzoyRz
Operator 2026-09-07 (evening): Claude sessions stay on notplants-orchestrator;
on the cc-ci host the orchestrator is an opencode agent (agents.toml:
opencode/glm-5.2) steered from https://oc.ci.commoninternet.net, next to the
upgrader/report sessions. claude-install and CLAUDE_BIN are gone from the
modules; launch-upgrader/report print the real UI URL; README "Operating the
orchestrator" rewritten and a "Weekly upgrade run" section added.
Secrets: only cc-ci's, under /secrets/files with runtime paths symlinked
(README §4 table), nothing from other projects — tangled key and the
tailscale line dropped from the host.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqkQq3CDmFWcQ7u1LzoyRz
The orchestrator's flake now builds the machine it shares with the cc-ci CI
server: `nixosConfigurations.cc-ci` composes cc-ci's nixosModules.cc-ci-server
(new flake input, nixpkgs + sops-nix follow ours), this repo's orchestrator
module (nix/modules/cc-ci.nix, exported as cc-ci-orchestrator, `cc-ci` kept
as an alias for notplants-nix) and the new nix/modules/orchestrator-host.nix
— the host contract those units always assumed (loops user, claude/opencode
CLIs, opencode web server + tailnet-only UI on 8443 since traefik owns
80/443, nix-ld, tool set, `ssh cc-ci` → loopback).
nix/hosts/cc-ci/{hardware,networking}.nix are PROVISIONAL copies of the old
server's layout so the flake evaluates; they get replaced by the
nixos-infect output of 195.201.88.249.
README.md is the deploy guide (Hetzner Debian → nixos-infect → this flake →
staging → data restore → cutover). archive/ holds the retired Incus/Hetzner
orchestrator host configs, the old terraform and the migration plans;
references updated. cc-ci-plan/plan-cc-ci-combined-host.md is the working
plan for the move.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FqkQq3CDmFWcQ7u1LzoyRz
The Gitea host was stored in the credentials env file, so a literal
secret scanner treated it as a secret and refused two READMEs in another
repo that link to git.autonomic.zone.
It was never secret: advisory-scan.py, pr-survey.py and recipe-report.py
already default it to git.autonomic.zone, and it is this repo's own
remote host. It is a bare hostname with no scheme, userinfo or path.
Removed from the env file and defaulted here the same way the Python
already does. GITEA_USERNAME and GITEA_PASSWORD keep their :? guard --
those are real credentials and must still fail loudly if unset.
Move plausible from weekly to external in used-recipes.md. cc-ci still
deploys/tests it (the recipe + tests stay enrolled, !testme bridge
unchanged), but it is no longer surveyed/upgraded by the weekly /upgrade-all
run — maintained elsewhere, like uptime-kuma.
The open plausible PR #5 (clickhouse 23.4.2.11 -> 23.4.6.25) from this week's
run stays open for the operator; no future weekly PRs will be opened for it.
The weekly /upgrade-all parent session and the /recipe-report session now
run on tinfoil/deepseek-v4-pro (LOOP_MODEL + REPORT_MODEL in upgrader.env).
The hourly supervisor stays on opencode-go/glm-5.2 (SUPERVISOR_MODEL default
in launch-supervisor.py, not overridden). Subagents already bind deepseek via
the cc-ci repo's opencode config (fix from 2026-08-10, verified this week:
all 16 subagents across both waves ran deepseek-v4-pro).
LOOP_TIER=zen is kept so the tier check passes; the watchdog's usage-limit
probe sends the deepseek model name to the zen endpoint, which returns 200
(not 429) → resume immediately — correct, since tinfoil has no rolling usage
limit to wait out. Verified the probe behaviour with a direct curl.
Root cause: the 2026-08-14 run stalled mid-recipe on 'Insufficient balance'
(opencode zen workspace balance exhausted), then sat unfinished for 40h while
the supervisor cron spun hourly unable to recover it. Deepseek (pay-per-use
API key) has no rolling balance limit, so this can't recur.
Also documents the session recovery in JOURNAL.md (the stalled run was
completed via a fresh scoped upgrader — the original 2.58M-token session was
unresumable: the inference endpoint silently drops the oversized request).
autonomic-bot
loops
autonomic-bot
nptest2
notplants
cc-ci-orchestrator