Bring the worldchunks drive back into use after it drops
On 2026-09-23 the USB earth-tile drive on k3m1 dropped for a second. systemd unmounted it and stopped nfs-server, which nfs-utils makes require every exported p...
SP14-T14 ruling — the MediaService ConfigMap guard and Step 2's full SP14 test run
SP12 final fixes — realtime STT measured as SP13 calls it (A-3 bound 9.0 s), recordings claimed before processing (B-5)
Rulings SP12-FF (probe language), (threads), (A-3), (claim), (B-5) in the SP12 spec §7; §4.4, §4.7, §7 items 7 and 20 and §10 updated; plan Task 18 A-3 and B-5 ...
E2's streamed usage chunk reports the metered usage, never zeros
Hugging Face, Google and Cohere text streams carry no usage event, so the assembled response counts 0 input and 0 output tokens. E2 built its stream_options.inc...
The load verdict's budget is the amended acceptance bound, 9.0 s
Task 18 A-3 asked for p95 <= 4.0 s with 4 realtime streams and a batch. k3m1's Haswell CPU cannot reach it with faster-whisper-small: with the language given (S...
SP05-T32 ruling — F1/F2 field typing, character counting, the unnamed file, auto as a source
F1 takes back the auto it answers as its source language
A caller that echoes F1's from:"auto" is translated with detection instead of refused, and F2's placeholder name says why it exists. Ruling SP05-T32 (e).
Translation and OCR façades on the router for the Tool Plane
F1 POST /api/ai/v1/translations {text, to, from?} and F2 POST /api/ai/v1/ocr (multipart file, optional filename), SP05 spec §4.17: the caller's session, SP04's ...
Claim a staged recording before processing it, so a long call is never stored twice
DaemonService hands ConversionService every row that is still unprocessed: on its poll tick, on every voip.recording.received message (which does not wait for a...
SP14-T13 fix round 1 ruling — a KM2 dial is claimed before kamoai-service is called
One call per AI per meeting — a dial is claimed before kamoai-service is asked
Two overlapping requests to dial the same AI into the same meeting (a double click, two organizers, a person and the auto-join sweep) could each pass the not-li...
Park the KamoUniverse game, holding its Scylla and Valkey at 0
The game is not in use or development, so it is offline to free its resources. Nothing is deleted: Scylla and Valkey keep their data on k1m1 hostPaths, the kamo...
Measure realtime load as SP13 calls it, and give each transcription 8 threads
The load probe's realtime sessions sent no language, so every turn paid Whisper's language detection (9.4 s alone at 3 threads on k3m1), a call shape no product...
Run llama-server with one thread per core of its CPU limit
Ollama starts llama-server without --threads, so it counted k3m1's 24 physical cores and ran 24 compute threads inside the container's 8-core quota. The contain...
SP05-T20-ship ruling — framework shipped as Task 16's head; step 9 in-pod, no port-forwards
SP14-T13 ruling — AI participant state machine, sweep retries and budget, I3 participant check
AI members join meetings — invite, auto-join at start, runtime state, companion context
SP12-T17 ruling — whisper-asr retired at ship time, not after a 24 h soak that could observe nothing
Redeploy f9a0aa1 — run 69's deploy job hit a Kubernetes API timeout
Run 69 built f9a0aa1, but its deploy job failed at 'kubectl create ns' with 'Timeout: request did not complete within requested timeout'. No code change.
SP04-F1 ruling — an AI member's own chat session routes, budgets and meters as actor AI
SP14-T12 ruling — M3–M5 answer JSON, keep ties on one page, tolerate foreign UUIDs, follow M1's invite rule
Meeting history, meeting detail with dial-in and transcripts, inviteable AI members
M3 GET /api/meet/ai-candidates, M4 GET /api/meet/meetings, M5 GET /api/meet/meetings/{id} (spec §4.4.1). Every answer is application/json; a page never splits m...
An AI member's own chat session routes, budgets and meters as actor AI
Member chat (/api/ai/sessions/**, /ws/ai-chat) routed every member as HUMAN, so an AI member's session (its SP11 workstation browser) reached UNVERIFIED models,...
SP04-F2 ruling — façade timeouts cover the whole failover (SP10 390 s, SP13 summary 190 s), master §11 row
Detailed transcription — timed segments, speakers, language and model provenance
POST **************** (SP14 spec §4.4.3) answers a recording as timed segments with the language and the provider and model that heard it, for MediaService's pe...
SP12-T17 rulings — delete whisper-asr before the push, scan Secrets by key, pin the retirement in a test
Retire it — every transcript now comes from Kamo Speech through AIService
Transcription runs on Kamo Speech, and PHI audio is routed only to in-cluster services
SP12-T16 ruling — no call traffic: Task 16 Step 5 becomes Task 17's precondition
Like what you see shipping?
All of it arrives in your workspace on its own. Start on the free plan and read this page again in a month.