Beam-2 with a temperature-fallback ladder made the final-model warmup
run all three temperature retries under beam search before the server
bound its port, so /health was refused for ~8 min and STT was down that
whole time on every roll (and hinted at slow per-utterance decodes).
Production now runs the accurate large-v3-turbo model greedily at a
single temperature, keeping the proper-noun priming prompt that fixes
names like Amy/Córdoba - fast startup, fast decodes, accuracy intact.
Beam stays env-tunable (HERMES_STT_FINAL_BEAM_SIZE) for a future pass;
serve-before-warmup is a recommended follow-up so cold start never
blocks readiness.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
A transcription that landed while the private Whisper service was
restarting (an image roll) crashed hermes_stt_client.py with a raw
urllib ConnectionRefused traceback that got dumped into the
conversation. The client now retries the request with backoff (up to 5
attempts, ~10s - long enough to ride an STT pod restart) and, on a
persistent outage, exits with one concise line ('speech transcription
unavailable...') instead of a stack trace. Delivered via the coordinator
ConfigMap; the next reconcile picks it up.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Proper nouns (Córdoba, Cancún) and dropped words came from decoding the
committed transcript with the small model. The image already ships
large-v3-turbo, so the final decode now uses it with beam_size=5, a
temperature fallback ladder, and a proper-noun/accents initial_prompt
that fixes first-pass capitalization and diacritics across EN/ES/RU;
the rolling previews stay on tiny at greedy so the on-the-fly feel is
unchanged. The accurate decode runs in the speculative predecode during
the end-of-speech silence and is cache-reused at commit, so perceived
latency stays low. All decode knobs are env-overridable for on-device
tuning (beam/temperature/prompt), with small as the guaranteed-present
rollback if turbo underperforms on the Jetson. 206 STT tests pass.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
The size cap was applied to the raw wire body, so a nodes list (huge
because of status.images) truncated before the image-stripping ran and
came back as a truncation notice. Accept up to 6 MiB on the wire to
parse and clean, then enforce the 384 KiB model-facing cap on the
stripped result - nodes now returns real data. Delivered via the
cluster-read ConfigMap; picked up on the next pod roll.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
The chat prompt still forbade all cluster access, so Hermes told users
it had no visibility even though cluster_read is live - it now knows it
has a read-only cluster tool (no Secrets/Vault) and should use it rather
than deny. Adds tool/research-conditioned acknowledgement guidance: when
a turn needs a lookup, plan or calculation, open with one short 'on it,
~ETA' line then deliver the full answer; simple questions get no
preamble. In spoken mode that first line is read aloud. Fixes two
pre-existing exact-match test pins the HUX/cluster rollout had grown
(plugins.enabled list; a fieldRef env comprehension).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Node .status.images (every cached image on the node) overflowed the
size cap and left json.loads parsing a truncated blob, so a nodes query
came back as a non-JSON error. The de-noise pass now summarizes that
list, and genuine truncation returns the readable prefix with a
narrow-your-query hint instead of an error.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
The read-only cluster_read tool (and the HUX-12 producer) reach the
Kubernetes API through the kubernetes Service, which kube-proxy DNATs
from the 10.43.0.1 ClusterIP to a control-plane node on 192.168.22.11-13
:6443 - addresses the tenant egress except-block was dropping, so calls
failed with connection-refused. Egress now allows the ClusterIP and
those three apiserver endpoints on 443/6443. RBAC still bounds what is
readable (no Secrets); nothing else about the isolation changes.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
worker.bstein.dev (the hermes-agent Deployment) gains the same HUX
shape as chat, staged and inert: a foundation-only hux sidecar on the
reviewed WebUI image line (Flux setters bound, 5s probe budgets), an
init that provisions the HMAC identity as slot-100 on the durable home
subtree (create-once context key, O_EXCL subject binding, per-pod
worker key; no relay/router/evidence keys so those trusts fail closed),
and observe-only hook env in the agent container with the runtime
plugin mounted but deliberately NOT enabled - activation is a reviewed
one-line flip per docs/hux/WORKER-PLAN.md, which carries the rollout,
verification gates, canary/rollback ladder and open questions.
Cross-surface continuity remains unclaimed until the live gates pass.
7 new topology-adaptive delivery gates green.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
RBAC: the built-in view ClusterRole (which never includes Secrets, so
Vault-managed material stays structurally invisible) plus a read-only
extra for nodes, namespaces, PVs, storage classes, CRDs, Flux resources
and metrics, bound to the chat service account. Tooling: a cluster-read
plugin registers a GET-only cluster_read tool against the in-cluster
API using the pod's projected token - secrets paths refused in the
handler as well, malformed segments rejected, responses bounded and
stripped of managedFields noise. Classified read_files/low in the HUX
capability map. RBAC applies on push; the tool activates when the pods
next roll (bundled with the round-3 voice build).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Real traffic showed the first enforcement pass blocking core assistant
faculties: skills listing and the sandboxed Python classified as
unknown external side effects, browsing denied by default, with no
approval surface in the chat flow. Enforcement returns to observe-only
fleet-wide while the approvals UX and default grants are reworked, and
the capability map now tells the truth about the real toolset: the
Python sandbox is internal shell work, skills/todo/clarify/vision are
reads, browsing is network (medium), image generation writes an
artifact through the trusted broker. Unknown tools remain fail-closed.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Removes the canary partition so all four tenants receive the reviewed
build-22 image with the HUX topology and enforcement (every lifecycle
gate passed on ordinal 3). Bumps the hux probe timeouts to 5s: the 2s
readiness exec budget timed out 405 times in 135 minutes on the loaded
ARM node — interpreter spawn cost, not service health — and the
resulting flapping stalled the StatefulSet roll at the canary. Converts
the HUX_IMAGE_TAG/DIGEST setters to block style with the marker on the
value scalar (Flux cannot attach setters inside flow mappings, which
left the metadata stale at build-21) and corrects them to build-22.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Park/resume and gateway-owned stop receipts passed the live gates, so
HUX_TOOL_ENFORCEMENT=1 ships to ordinal 3 (partition unchanged). Every
canary tool call now requires a HUX release; the fleet stays observe-off
until the approvals UX proves out under real traffic.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
hux.autonomy joins the canary flags (which also unlocks multimodal via
its dependency chain). Enforcement remains 0: approvals, gates, budgets
and stop receipts run observe-only until park/resume passes live.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Artifacts, research, friendly modes, multimodal and onboarding join the
canary flag set on ordinal 3 (partition unchanged, enforcement still 0).
Autonomy and release follow-through remain off pending their gates.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Ordinal 3 only (partition unchanged): activity timeline, projects,
privacy and memory control join the foundation flag, observe-only.
The roll also serves as the HUX-11 restart-persistence gate via the
recorded identity hashes.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Re-applies the staged HUX topology pinned to the reviewed build-21
image (git-2f535d3a...-build-21-release@sha256:e5b9b2fa...), with the
first-activation posture: HUX_FLAGS=hux.foundation only,
HUX_TOOL_ENFORCEMENT=0, and a RollingUpdate partition of 3 so only
hermes-chat-tenant-3 rolls. Adds the /healthz auth bypass on the chat
proxy so HUX-12 health receipts can observe a real 200, points the
evidence policy at it, and makes the delivery flag gate progressive
(foundation first, cards enabled per lifecycle acceptance; unknown
flags still never ship).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Restores the five services/hermes manifests to the live state and
removes the evidence policy/RBAC resources, so pushing this source
chain applies nothing to the cluster beyond the inert hux_mode.py key
in the auto-router ConfigMap (old pods are protected by the plugin's
degrade-to-noop import shim). The activation topology returns as a
dedicated commit pinned to the newly built WebUI digest, per the staged
release sequence in docs/hux/HANDOFF.md.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
The generator now ships the HUX-06 adoption module alongside the
plugin, and the plugin degrades to its previous behaviour (no mode
adoption) if it starts against a ConfigMap rendered before the module
existed - the plugin can never fail to load mid-transition.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Activation-layer staging, fail-closed until enablement: a per-tenant
hux-evidence-producer sidecar on the exact reviewed WebUI image runs
hux_producer.run_once on a 60s loop, inert until the Vault-staged
evidence key (tolerant init, tmpfs, 0400, staged only for the hux
service and producer containers - never hermes or webui), the policy
ConfigMap, and the scope ConfigMap exist. Adds least-privilege
read-only RBAC (pods+statefulset in hermes, the single named Flux
Kustomization), tenant egress to the Kubernetes API ClusterIP and the
traefik edge, the policy allowlist, hux_producer packaging in the WebUI
image, and a third expected WebUI consumer in the Flux release
renderer. Delivery and image-automation gates enforce the boundary.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
HUX-04/HUX-08: after a successful tool execution the hux-runtime plugin
now registers freshly written files as typed artifacts (create or
immutable version by conversation-scoped title, bounded to 1 MiB,
deterministic idempotency keys) and records up to three deduplicated
web sources from real network/web_search output. Emitters are fail-open
and can never break the tool result. Delivered through the plugin
ConfigMap; 95% branch coverage; quality contract updated.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
HUX-06: the auto-router now consults the loopback HUX service (worker
trust, HMAC-derived conversation id from the persisted session) for the
conversation's selected friendly mode and maps it onto the existing
provider-neutral pools: fast->auto/fast, thoughtful/research->auto/deep,
create->auto/balanced; private pins the local route and refuses hosted
overrides. Explicit UI picks keep precedence; every HUX absence or
failure falls through to the previous behaviour unchanged. 95% branch
coverage; quality contract registers the new module.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
Defense in depth: the tenant router deletes every browser-supplied
X-Hux-* header before asserting its own HUX identity headers, so no
client can forge subject, trust class, or relay key. Regression covers
X-Hux-Subject, X-Hux-Trust and X-Hux-Relay-Key forgeries.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BvMSXH8VH2tMWXanb8SJdf
