make quality-high reformatted three expressions this branch added today: the chat-target chain in the translate role's vars, an assertEqual in the gateway engine tests, and the wanted comprehension in the lane probe. No logic and no rendered value changes; the YAML break folds to a space inside a double-quoted scalar.
Kept separate from the changes that caused it, because a reflow mixed into a behaviour commit hides the behaviour.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The oracle asks a chat model whether a translation states its source. llama3.2:1b cannot answer that: on the suite's own fixture it scored an unrelated text at the middle level instead of the bottom one, and asked as a yes-or-no question it called two identical texts different. With it the suite reached no verdict at all, reporting that the judge read none of the 47 languages the catalogs hold machine translations for. Qwen2.5-3B answers both correctly, and the same run then judged 248 samples.
The pin lives in the lane inventory rather than in AI_LOCAL_MODELS, because a judge that can separate a faithful rendering from an unrelated text is a demand no deployment makes, and the default should not grow a gigabyte to satisfy a check. The litellm config builds its routes from the merged application config, so a lane override reaches the model list and its window without touching the repository default.
LANE_MODEL_ALIAS repeats the one name in utils/inventory/tools.py, because ensure() asks whether the lane already serves before the deploy that writes the rendered inventory, so the probe cannot read it from there. A lane serving one model while the probe expects another never reports ready and redeploys on every call, so a test renders the template and holds the two names equal.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The lane ran litellm but never republished its port, so a host-side probe hit the compose proxy on the same number and read 502. The lane could therefore never report ready, and the oracle suite skipped itself to a green exit instead of asking a judge. A handler reads the port from the role the way translate.py already does, the runner forwards it, and TOOLS_LANE_KEYS takes it up: compose_file_args layers the override in without checking its variables, so an .env written before the key existed would abort every compose command rather than merely drop the lane.
The gateway's master key is pinned in the lane inventory for the reason the System One key already is: the provisioner would otherwise generate it into a vault the checks cannot read back, and the oracle judge is not a deployed consumer with a virtual key of its own.
serving() stops treating reachability as readiness. A lane brought up before AI_LOCAL_MODELS named a different model keeps serving the old one, and a gateway holding a key these checks cannot use looks healthy from outside while refusing every call they make. It now asks /models with the bearer token and requires the configured alias, and requires the gateway to offer every language the catalogs name, because that advertisement is what the translation run waits on. Without this the lane waved itself through and no change to its roles ever shipped.
Two lane-scoped overrides come with it. The ollama role caps itself at four CPUs, which left the host at 361 percent while twenty cores idled, so the lane gives it the host count the way the engine entry above it already does. And the lane pins the router to its declared order: the decider may send a pair the engine serves to the chat model, which a batch over 183 catalogs cannot pay for. Deployments keep the decider.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Three faults, one mechanism each, that together let the oracle suite report
a verdict it never reached.
The address. lane_url handed out the runner's compose service name, but the
stack's containers resolve through the deployment's CoreDNS, which never
learns one: getent hosts tools answers nothing inside them, so every check
that asked the lane from there skipped itself. The bring-up now writes the
runner's address, picked as the one inside INFINITO_SUBNET because the
runner sits on every network it was ever brought up on, and the checks read
it from there. It is rewritten on both paths, because a restart gives the
runner a new one.
The readiness. serving judged the engine alone, so a lane whose engine make
i18n had already brought up counted as up while System One was down. It now
probes every role the caller names, which is why the runner republishes
System One's port: a container IP is unreachable from the host that decides
whether to deploy.
The verdict. A silent lane made the suite skip, and a skip exits green.
Only a deploy that is already running leaves the stack alone and exits 0
now; a crashed deploy and a deployed-but-silent role both exit 1.
The override's third :? guard is added to the gate that decides whether the
file is layered in at all, next to the one that was already missing. Without
it an .env written before these keys aborts every compose command, down to
compose-down, instead of dropping the lane. A test now keeps that list and
the override's guards in step.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`ensure` deployed unconditionally, and the lane is pinned to
`loading_model: task`, so every `make test-oracle` reinstalled the whole
model selection and restarted the engine onto it. Callers want a lane that
answers, not one that was just rebuilt. Three prefetches were found stacked
inside the engine, one per deploy, contending over the same package
directory so that none of them finished.
Reachability is the wrong question: the engine answers as soon as it has
booted, carrying whatever subset of models it found, which is the state a
run then waits on forever. What it reports is compared against what the
lane pins instead.
The comparison reads the language subtag, because the lane pins `zh` and
the engine calls the same language `zh-Hans`; comparing the codes as given
never matches and would have left the check permanently false.
Only the engine is judged. It is the lane service the runner republishes to
the host and the expensive one to rebuild, the lane deploys as a unit, and
a sibling that is down reports itself: each check names the address that
failed to answer.
The two messages flush, or the line saying which branch ran arrives after
the work it describes.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The lane is no longer LibreTranslate-only, so `i18n` is the wrong name for
the runner, its compose override, its inventory and its env keys. Renames
to `tools` throughout and points the lane at
`svc-ai-libretranslate-engine`, which is the role that actually owns the
`libretranslate-engine` service key.
The six cache host paths were each written out in full. They now derive
from `DIR_CACHE`, one plain string in `group_vars/all/05_paths.yml`, via a
`cache_paths` handler; the generated values are byte-identical. The handler
is registered directly after `cache_stack` rather than late, because
`registry_cache_max_size` reads `INFINITO_CACHE_REGISTRY_HOST_PATH` and
returns early when it is empty, which silently dropped
`INFINITO_CACHE_REGISTRY_MAX_SIZE` from the generated `.env` and broke
every `docker compose` invocation that layers `cache.override.yml`.
`INFINITO_TOOLS_MODELS_HOST_PATH` derives from the same base, so the model
bind the lane mounts follows the cache layout instead of repeating it.
The inventory pins `DOCKER_BIND_HOST=0.0.0.0`: the group_vars default keys
off `DOCKER_IN_CONTAINER`, which evaluates false for the inner deploy and
would leave the lane's services on the runner's loopback.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
kevinveenbirkenbachandClaude Opus 5