The kuma service_down nudge cannot name the service. mavpoll folds the whole
monitor_status gauge into one boolean fact keyed `service_down`, and the
phraser names a service only when the fact key is the service name, so the
message is always the generic "a service on homesrv is down". Every fifteen
minutes, with nothing to act on. A fact per monitor is the real fix and it is
filed as Vikunja #444; this is what to do until then.
`disabled_rules` in mavend.json subtracts from loop.DefaultRules by name.
Config only subtracts — rules stay code, the set stays canonical and ordered
as written. A disabled rule is not gathered for either, since the gatherer
derives its key set from the rules it was given. Unknown names are ignored so
deleting a rule cannot brick a config that still lists it, and the boot log
says what was dropped, because a rule that vanishes silently looks exactly
like a rule that is broken.
Deploy turns service_down off.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
mavend seals its encrypted database in `defer st.Close()` when run() returns.
It had not returned since 2026-07-21. Every restart since then decrypted the
same eleven-day-old ciphertext and rolled back everything written in between:
the Telegram nudge that kept firing was a fact being un-written on each boot.
The goroutine dump named it. main → srv.Close() → ipc.(*Server).Close →
wg.Wait(), waiting on per-connection goroutines parked in readFrame. Close
shut the listener and nothing else, so the idle persistent sockets held by
mavweb, mavpoll, mavcaldav and mavmaild blocked shutdown forever. `docker
compose stop -t 60` spent the whole sixty seconds and then took a SIGKILL.
So: track the accepted conns and close them, in ipc and in voice, which had
the identical defect. Bound all three waits — the two per-server ones and the
worker wait in main — because the seal matters more than any single in-flight
call. A dropped RPC costs one reply; a missed seal costs a session.
The regression test leaves a client connected and idle, which is the case the
old tests avoided by closing the client first.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
Two things that were each half-done.
The prompts handed the model copyable examples. chatSystemPrompt lost its
openers this morning and the "Я подумала, что" tic went with them, but "не
забыл ли я" appeared in its place: the removed example had been suppressing
the masculine self-reference by accident. Two predicatives are not enough
signal, so the rule is now stated as morphology (-ла) rather than as a pair of
words — a suffix rule generalises where an example only gets copied.
querySystemPrompt had the same defect and gets the same treatment; its "вот что
я нашла: " opener is deliberate and stays.
internal/kiwix had no caller. actions_query.go said "once internal/kiwix is
wired into this chain" and that never happened. It is wired now, between the
notes pass and the web source: everything of his answers first, and only what
is left over is looked up. Off unless a `kiwix` block names a server and a book.
Reading the search snippet does not work. Kiwix builds it from wherever the
keyword matched, which on Wikipedia is the navigation box at the foot of the
page — the first version of this answered "что такое фотосинтез?" by reciting
"Ecological economics Ecological footprint Ecological forecasting …". Client
grows an Article method; the head of the article is the lead paragraph, which
is the definition the snippet was meant to be. Verified on the box: the same
question now answers correctly off the ZIM.
Only the rewritten query leaves the process. A test asserts it: a turn carrying
a stored note must not put that note in the search string.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
The grammar examples in chatSystemPrompt were full clauses — ("я подумала",
"я рада") for her, ("ты сказал", "ты забыл") for him. A 1.7B copies those
instead of generalising from them.
Observed on homesrv 2026-08-01, in one session: all three chat replies
opened with "Я подумала, что ...", and one ended "...немного тревожусь.
ты сказал" — the second example pasted onto a finished sentence, which
reads as a truncation and is not one.
Contrastive pairs replace the openers, so the rule reads as a correction
rather than a template. The him-examples are dropped; the "ты" instruction
carries that on its own, and those two produced the worst output. A closing
line tells her not to echo the instructions, because a small model treats a
quoted string as licence to reuse it.
Verified after rebuild: three chat turns, no "Я подумала" opener, no
dangling example, feminine forms intact.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
Direct egress to api.telegram.org does not work from homesrv, so every
away-channel send timed out and the service_down nudge retried once a
minute forever. The sink already had a Proxy field wired to
http.Transport.Proxy; nothing had ever set it.
The relay is the x-ui socks inbound on the host, port 10808, addressed from
the container as the maven_default bridge gateway. That also needs a ufw
rule, because the bridge subnet is not otherwise allowed to reach a host
port and the SYN is dropped rather than refused. The rule is recorded in
the config next to the address, since the address alone is not enough to
reproduce this on another box.
Verified: five minutes after restart, zero send errors where there was
previously one per minute.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
net/http wraps every transport failure in *url.Error, whose Error() prints
the request URL. Telegram accepts the bot token nowhere but the URL path, so
a send failure wrote the live token into the daemon log. On 2026-08-01
homesrv could not reach api.telegram.org and did that once a minute for as
long as the network stayed down. The token lives in deploy/telegram.env to
stay out of the repo; putting it in `docker compose logs` undoes that.
Both error sites now go through redact. The structural branch rewrites
url.Error.URL and keeps the type, so errors.As still matches; anything else
falls back to scrubbing the rendered message. No minimum-token-length guard:
a one-character token would shred the message, but that beats leaking it.
DESIGN.md still said the classifier cascade was the path that runs today
with llmrouter wired nil, and gave the resident checkpoint as Qwen3.5-0.8B.
Both stopped being true on 2026-07-31.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
7ab9b48 committed models/stt and models/tts as symlinks to their own
absolute paths:
models/stt -> /home/kami/apps/Maven/models/stt
They came from an agent worktree under scratchpad/wt, where a link back
to the main checkout resolves. In the main checkout it points at itself.
Both paths are gitignored, so checking out that commit overwrites the
real model directories without warning and git says nothing. On homesrv
it destroyed models/stt/ggml-small.bin and the piper voice, and mavsttd
crash-looped on the missing whisper model.
The directories are host state fetched separately, per deploy/README.md.
Nothing under models/stt or models/tts belongs in git.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
One linear chain of 35 PRs, reviewed and fixed. The eleven fix branches were merged onto fix/integrated and fast-forwarded onto this tip, so the review findings land as commits here rather than on the individual PRs.
make test and make build pass.
The vendored copy of github.com/kami/hexis predated Client.WithToken:
no token field, no setter, no header hook, and an unexported httpClient,
so there was no way to attach auth from outside the package. wireEcosystem
handled that by refusing to wire Hexis at all when a token was configured,
which was the honest reading of the code but left the deployment silently
without its executing service.
go.mod already replaces the module with /home/kami/apps/hexis, and that
source has had WithToken and the Bearer header for a while. Only the
checked-in vendor/ copy was stale. Refreshed it (client.go plus the new
capability.go) and wired Hexis like Nexus and Praxis.
Two tests cover the outcome the refusal was standing in for: a configured
token reaches the wire as Authorization, and no token still wires unauthed,
because Hexis without auth is a valid deployment on a trusted box.
Also corrected the discoverCapabilities comment. It claimed the client
stamped the correlation header on Execute only; do() stamps it on every
request, and did before the re-vendor too.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
IngestMail was AuthRead and SetTaskStatus was AuthWrite, and they answer
the same question: may this module change what is on his lists? The old
argument for AuthRead — ingestion is additive, it can only produce
candidate tasks — is still true and is the weaker half, because a
compromised mail reader that can fill the review page indefinitely is
not a read.
Nothing loses access. AuthWrite outside WriteFact only requires
enrollment, which mavmaild already has, and the method does not exist
unless the operator wired a mail block.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
Two fix branches independently added a Gate to internal/llm. One is priority
between a voice turn and a background job, the other is admission control while
the resident model is swapped. They are orthogonal and both are needed, so the
swap one is now SwapGate, with SetSwapGate to install it.
Complete takes the priority gate first and the drain second. A background
request can wait a long time on priority, and counting it as in flight against
the drain that whole time would stall a swap on a request that has not started.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
capture_stop held the IPC request open for the whole map reduce, up to
twenty minutes. A voice turn that says "хватит" waited for forty model
calls before Maven said anything. Stop now returns the transcript and the
summary runs on a goroutine in the daemon's WaitGroup, on the daemon
context so a client that hung up does not cancel the only readable record
of the meeting.
With no summary and save_transcript false, writeNotes wrote nothing at
all: an hour of meeting left a blob that prunes in seven days and no
trace in the note store. The transcript is written instead when the
summary is missing. That flag decides whether the verbatim record is kept
in addition to a summary, not whether the meeting is remembered.
The wire carries the session token now, and the contract comments say
what the code does: the summary is usually absent from the stop
response, and re running a stored blob is a manual job because no method
takes a blob id. The save_transcript comment says the cost is recall
corpus rather than disk.
Found in review of #73.
Four invariants the comments claimed and the code did not hold.
The recording lived in mavend's heap as one growing []byte, doubled at
Stop when the WAV was built. Frames now go to a spool file and the
transcript is read back off disk one window at a time, so memory is flat
whatever the length.
A store failure returned before transcription ran, so a meeting over the
blob cap produced no transcript, no summary and no note. It now records
the failure and keeps going, and the spool file survives until the words
have been read off it.
The session had no owner. Any module on the write rung could call stop on
a recording it did not start and receive the verbatim words of everyone
in the room. Start hands back a token and append, stop and abort require
it.
The duration cap was only checked when a frame arrived, so a phone whose
tab was closed left the slot occupied and every later start answered
ErrBusy with a meeting from last week. The wall clock is checked in
start, status, append and stop.
Smaller things in the same pass. Append compares the frame format against
the session format, so a client that switches sample rate mid meeting no
longer has its frames concatenated under a header that lies. One failed
STT window leaves a marker instead of discarding the other twenty four.
Summarize is separate from Stop and assigns the salvaged per chunk text
before it reports the error.
Found in review of #73.
Put takes a []byte, so storing a recording meant the whole recording in
memory. A two hour meeting at 16 kHz mono is about 230 MB of WAV, and
building it from PCM held a second copy of the same size in the process
that also owns the database and the resident model. PutFile stats the
file, hashes it in a stream and renames it into place, so the peak is one
buffer regardless of length. SpoolFile hands out the scratch file it
moves from, under the media dir so it shares the same disk and the same
permissions.
Audio also gets its own per blob cap of 512 MiB. The image cap of 64 MiB
is 35 minutes of audio, which contradicted the two hour session cap: the
long meeting was exactly the one that failed to store.
audio.WAVHeader is split out of WAVFromPCM because a spooled capture
writes a placeholder header first and stamps the real length at the end.
Found in review of #73.
The docker-shaped update block in TestUpdateBlockValidatedAtStartup has
source_dir equal to install_dir and no source_rollback, which is exactly the
deployment the new validation refuses. The fixture is meant to be the good
case, so it now says how the source is rolled back.
Found in review of #69.
The on-demand crawler was built over allow_hosts plus every watched host.
webfetch reads a non-empty allow list as these and nothing else, so a config
with one watch and no allow_hosts at all silently narrowed on-demand reading
to the watched site. Every other url he pasted came back as a flat refusal
with nothing in the log to explain it. The two crawlers now take two host
lists from one crawlHosts helper.
Crawl-delay was parsed into Rules and never read. The only pacing was the
fetcher's flat one request per host per second, which cannot express what a
site asked for, and deploy/README claimed the field was honoured. Page now
waits it out between the robots fetch and the page fetch, and a delay longer
than the turn fails the read instead of hanging it.
A robots.txt that failed was treated as no rules, so a site whose server was
having a bad minute became a site with no restrictions. A 5xx now refuses the
crawl. A 404 still means unrestricted, which is what the standard says.
The refusal check matched substrings of webfetch's message text from a package
that cannot import webfetch, so a reworded error would have silently turned
into a robots verdict. internal/crawl now exports ErrFetchRefused and
ErrFetchStatus and the adapter in cmd/mavend maps the webfetch sentinels onto
them. Robots group selection picks the longest matching agent prefix instead
of the first one in file order.
queryWeb passed a claim it could not serve when no crawler was configured, so
an unconfigured deployment answered a web question with an apology instead of
falling through to the model.
Found in review of #67.
Nothing at read time told a feed item or a crawled page apart from his own
notes. QueryNotes ranked every note by cosine and the notes answer handed the
nearest five to the phraser, so "что я говорил про переезд" could be answered
out of a stranger's web page, prefixed with "вот что я нашла: ". Recall now
excludes the read sources, rss: and crawl:, and the list is one place.
The feed answer needed a different read as a result, and it needed one anyway:
it scanned the last 200 notes of any source, so a busy day of voice notes pushed
the newest headline out of the window and she said "в лентах пока ничего
нового" while the poller was working fine. RecentNotesFromSource asks for feed
notes by source, so the window holds 200 of them.
Found in review of #66 and #67.
expect_not_called could pass on a step that made the forbidden call. callPaths
concatenates per server and callCount was a total, so slicing the concatenated
list by the total examined the wrong window. With praxis on three requests and
nexus on one, a fourth praxis call landed at index three and paths[4:] never
saw it, while the stale nexus call was reported as new. The mark is now
per server and the paths are taken per server from it.
Neither scenario ever produced an act, so all three fakes saw zero requests and
the fault lever changed no outcome. The two headline capabilities of the
harness had no coverage. act_degraded scripts an act against an enabled
allowlist row and runs it healthy, at 503 and healthy again, asserting the
reply, the call, the absence of a call on a tick, and that nothing was pushed
at him either way. That needed two seams the world did not have: allowlist rows
from the scenario, and a matcher on the real store rather than a nil API, which
would have panicked the moment any scenario produced an act.
expect_no_events compared bus.Len(), which stops growing at the ring capacity,
so a scenario long enough to fill the ring made every later expect_no_events
pass unconditionally. It counts publishes through a subscriber now.
A scenario could not express a fact below full confidence, because write
hardcoded 1.0, and morning_missed annotated its ambient step as if it could.
factPriority branches on exactly that, so no replay could reach the low branch.
signalStep takes a confidence, the ambient step sets the 0.6 the ambient path
writes, and the event line carries the priority so a scenario can assert it.
TestSimulatorIsDeterministic compared the transcript against time.Now, which
fails for the half hour a day the scenario itself covers. It checks that every
stamped line falls inside the scenario span instead. TestSimulatorRefusesBackwardsSteps
tested the forwards case, because reaching the backwards branch ended the test.
A fatalf seam makes the refusal observable.
Smaller notes: the step doc comment now states which assertions are run scoped
and which are step scoped, audioText parses the golden manifest once per world
rather than once per step, and the feminine checks list the masculine form with
its following character, since the earlier check on a comma alone passed on
"записал что ты выпил воды".
Found in review of #79.
Saving a description writes recall corpus. writeNote embeds it under
media:image:<id>, a source no enrollment owns, and the method sits at
AuthRead, so any enrolled module could put a small VLM's guess into what
Maven knows and have it come back in a later turn as something she
believes. The describing half stays a read; save_note is now held to the
same source-scope rule WriteFact is, and the stored text carries a
marker saying it came off a picture.
Three doc comments said the method exists only when vision is enabled
and the code says otherwise. The code is right, and storing without
describing is the state this box is in, so the comments were corrected
rather than the behaviour. A request carrying both data and id used to
take the id branch and drop the bytes without a word; it is refused.
A media dir that cannot be created and a vision endpoint that is a typo
were logged at wiring time and the capability just stayed off, which is
the hardest kind of misconfiguration to notice. Both fail at startup.
runPrune was the one loop started with a bare go and not in the daemon's
WaitGroup, so shutdown did not wait for a prune that was deleting files.
Found in review of #72.
checkPrivate validates the configured endpoint literal and validated
nothing after it. The client followed redirects, so a 302 from the local
llama-server would have sent the photo, as a data URI in the POST body,
to whatever the redirect named. "No provider in this repo may upload a
blob" was true only of the first hop. Redirects are refused now, and the
reply is read through a cap rather than however much the endpoint feels
like sending.
ValidateEndpoint exports the same check so config can fail at startup on
a typo instead of logging once and leaving vision quietly off.
Found in review of #72.
"что нового?" is a greeting, and the feed matcher claimed it: "нового" was a
feed noun and "что" an ask. With no feeds block, which is what ships, the answer
to hello was "я пока не читаю ленты — они не настроены". A newness word now
needs a named topic or a real feed noun beside it. The topic prepositions lose
"о" for the same class of reason: one rune of filler produced a category of
whatever followed it, and then "по этой теме в лентах пока ничего".
An undated feed was re-noted in full on every boot. Dated items are deduped
against the durable mark, undated ones against a map that dies with the process,
so five items became five more on the next start, stamped now, at the top of the
recent-notes window. A crash loop made that a flood. The mark is now set for an
undated feed too, and its existence marks the first poll after a restart as a
resync: those items are recorded as seen rather than written.
A burst larger than max_items lost its middle. The poll walked the feed
newest-first, stopped at the cap, and marked the newest item written, which put
everything below the cap behind the mark forever. The cap now applies to the
oldest candidates and the mark follows what was written, so max_items paces
instead of dropping.
The category tag was read out loud: "Заголовок [технологии]" went through piper
brackets and all, because the answer path took the whole first line. The tag is
parsed off for reading and is now the only thing a topic is matched against.
Matching the whole note meant "что нового про погоду" hit any tech headline
whose link contained "pogod".
Also: the charset comment on dec.Strict described something Strict does not do,
and a skipped feed is named in the log.
Found in review of #66.
Put wrote the blob and then the sidecar. A full disk or a crash between
the two left bytes on disk with no sidecar, and List walks sidecars, so
Prune could never see them: Put returned an error and an image nobody
knew about became permanent. The sidecar goes first, a failed write is
rolled back, and Prune also collects blob files that have no readable
sidecar and are past retention, which picks up whatever an older build
leaked.
The per-blob cap bounds one call and nothing bounded their sum. Content
addressing does not help, because one flipped pixel is a different
digest, so 64 MiB per call and an unlimited number of calls fills the
disk mavend's database lives on. The store now carries a whole-store
budget, seeded from disk at open so a restart does not begin at zero.
Found in review of #72.
The only cap was 64 MiB of input, and a decode bomb is a small file. A
20000x20000 PNG of flat colour compresses to a few hundred kilobytes,
decodes to 400 million pixels, and flattenAndScale then allocated a
second buffer of the same dimensions before scaling anything. That is
3.2 GB of live heap from one request, on a laptop, in the process that
owns the database and the socket, and max_dim never got a chance to
help. The header is read first now and a source over forty megapixels is
refused. The scaler reads the source through At and allocates only the
destination, so flattening no longer doubles the peak.
Found in review of #72.
The fix pass ran in git worktrees, which need models/ symlinked in from the
main checkout to build. .gitignore covered the embedder and llm symlinks but
not stt and tts, so those two were committed as absolute-path symlinks.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TrVSBKe3RFDF4fGYKWYQnX
Nothing reads the microphone while Send is in flight, so the audio piles
up in arecord's pipe and arrives in a burst the moment dispatch returns.
A round-trip is p50 2.7s through the LLM router, which is about 90
frames of room, of him finishing his sentence, of the television.
The old code reset the VAD on the reply path only, and for a reason that
was not true: the comment said the VAD had been accumulating during the
round-trip, when its state is exactly what Feed left it as. The two
paths with no reset are the ones that mattered, because neither starts
playback and so neither is covered by the half-duplex gate. A text-only
turn fed the whole backlog into the VAD, and a Send error did the same
on every failed turn, so a dead socket drove a retry loop off backlog
alone.
The backlog was scored for barge-in too. Five frames delivered in
microseconds cut her off with audio recorded before she started
speaking, which is the opposite of what the five-frame guard is for.
Both are fixed by the same mechanism: measure the wall time the
round-trip took, convert it to frames, and discard that many before
anything looks at them.
Barge-in also threw away the 150ms that proved he was talking. The VAD
started from the next frame, so the first word of a short interruption
was clipped before whisper saw it. Those frames are kept in a small ring
and replayed after the reset.
A stuck aplay was worse than before this feature existed. Playing()
gates all capture, so a wedged child made her deaf rather than silent,
for the full 30s ceiling inherited from the fire-and-forget version. The
mute window is bounded by the reply's own duration plus a margin now.
Three smaller ones. "-barge-in -barge-in-rms 0" logged "barge-in on" and
then did nothing. The sent counter incremented before the error check,
so failed round-trips counted as shipped. And the threshold the operator
has to guess is now reported: mavwaked logs the mean energy of the
frames it suppressed while speaking, so he can set it from data.
Found in review of #76.
The day total rolls over at midnight and the poller had nothing to write until
the first spend of the new day, so at 09:00 the latest money_today fact was
yesterday's spending and looked perfectly fresh. The value now carries the
first instant of the window it covers, and a today question that the stored
window does not cover is refused rather than answered with yesterday's number.
Staleness was measured off the fact timestamp, which only moved when the figure
moved, so a quiet month was reported as data from three days ago while being
current. The value now carries when it was last read and the poller writes on
every read.
Amounts in an instrument the window diff never named were spoken with a numeric
instrument id as the currency. Instruments are resolved from one cursor-zero
diff, cached for the process, and an amount still unnamed is dropped from
speech rather than recited wrongly. "сколько я потратил вчера" was answered
with the month total, a real number to a different question, and is now
refused by naming the two windows she keeps. Income questions led with the
spending.
Found in review of #62.
dayDelta truncated both instants to a UTC day. A task due at 02:00 Moscow time
tonight read as due tomorrow, and one due at 23:00 last night read as due
today, so the two classes that decide the whole order were assigned from the
wrong calendar. Both sides are now truncated in now's location. Dated work also
lost to age alone because the later-due score sat below the age cap, and the
tail said "и ещё 3" with no noun and no Russian plural agreement.
The page hardcoded time.Now, so none of this was testable from a fixed clock.
It now takes an injectable clock, parses the due date in that clock's location,
parses ids and weights with strconv instead of a hand-rolled scan, caps the
resolved table and says so, shows who resolved each row, and reports a
promotion as the confirmation it is.
Found in review of #61.
A task extracted from mail deduped on the live-norm index only, so once he
finished it the row left the live set and the next poll of the same immutable
message re-extracted it as a fresh candidate. mavmaild is a read-only reader
and marks nothing read, so that repeats forever. Derived rows now carry an
ext_id built from the message uid and the extracted span, unique across every
status, while voice keeps live-only norm dedupe because saying an errand again
is the recurrence signal. A derived source can no longer capture straight to
open, and saying a task out loud that Maven had only proposed promotes the
candidate instead of answering that it is already in the list.
SetTaskStatus was classified AuthRead. Resolving a task is not additive, it
erases work off his list, so it is a write, and the row now records the caller
that moved it. ListTasks was unbounded. The list-query matcher claimed any
utterance with "что мне делать", including "с чем мне помочь", and the urgency
stripper matched inside words.
Found in review of #60.
The file comment named four regressions caught here. Three were not.
Nothing on this path resamples, because PCMFromWAV refuses anything that
is not already 16 kHz mono s16. Nothing exercises language selection,
because the hint comes out of the manifest already correct. And a bad
model path was the one condition that made the whole test vanish behind
a skip nobody reads. The comment now claims the two things that are
real, an explicitly set MAVEN_WHISPER_MODEL that does not exist is a
failure, and a missing fixture is a failure rather than a skip.
looseWordMatch accepted a different word. Four retained runes of "воды"
is "вод", so whisper hearing "выпил водки" satisfied the ru_fact
keyword, and "dis" let display, distance and discuss all stand in for
"disk". A case ending adds a rune, not a syllable, so the hypothesis is
capped in length as well as matched on prefix.
The spoken text lived in the generator and in the manifest with nothing
tying them together. Editing one left the other describing audio that no
longer existed, and at a flat ceiling of 0.34 over a five-word reference
a one-word drift passed silently. The script reads text out of the
manifest now, and the ceilings are set just above what each case really
measures against ggml-small, with the measurement recorded beside them.
Also: the test carried its own copy of the PCM to float32 conversion, so
a regression in the daemon's copy left the silence-gate assertion green,
and the manifest was validated for keywords but not for text, where an
empty reference makes every hypothesis score a WER of 1.
Found in review of #75.
The drain counted only the phrasing paths in internal/phraser. The router, the
replier, the mail extractor and the memory evaluator reach llama-server through
llm.Client, so quiesce could report zero requests in flight while the router was
mid-generation, and the old server was killed under it. The turn then finished
on the new model, which is the split turn the swap exists to prevent. llm.Client
now enters an optional Gate before every completion and LLMPhraser implements
it, so one counter covers every holder of the base URL.
A total failure also reported itself as a rollback. Swap set RolledBack on the
path where the rollback failed too, so the page rendered "rolled back to — she
is still answering, with the old model" over an empty model name and a daemon
with no model at all. The total failure has its own flag now, LiveModel stops
naming a gguf that is not loaded, and the log says another attempt can recover
without a restart, which is true.
The swap also ran on the connection every other page shares. ipc.Client holds
its mutex for a whole roundtrip with no read deadline on either side, so a load
froze /dash, /history and /notifications for minutes. mavweb dials a second
connection for /models alone. POST /models joins the route table, and the load
settings no longer come off a form that renders no input for them.
Found in review of #68.
Several assertions passed against code with the behaviour removed. The
independent-outage test shared no state to begin with, the capability
fixture used to prove read-only filtering was already mutating, and
route-level faults were simulated with a separate fake instead of the
shared one. The harness now takes per-route faults and a ticking clock,
so durations are measurable and one dead endpoint can be shown not to
mute a whole service. New cases cover a resolved reference with no
entity, a rejected credential, a malformed Praxis body, foreign items
in a scoped response, named truncation, traces staying out of facts,
and enrichment making progress while its oldest batch is backed off.
Found in review of #82.
The worker took the oldest pending facts by id and attempted them. Once
the oldest batch entered backoff the worker kept selecting the same
rows, found none of them due, and did nothing. One unresolvable fact
at the head of the queue froze enrichment for every fact behind it, up
to the hour-long backoff cap, forever. The worker now scans up to a
thousand pending rows and attempts the first batch that is actually
due. Retry state for rows that left the queue is forgotten, a failed
store write backs off the same way a failed resolve does, and the
status counts pending, backed off and exhausted over the rows it saw.
Found in review of #83.
The act path minted IDs per hop and trusted whatever Praxis returned
for a scoped attention query. A service that ignored the entity filter
would have had its unrelated items read back to the owner as his. The
handler now assigns one correlation ID at the top of the action and
passes it down, and drops any item the response did not tag with the
requested entity. Traces are written to the trace table with the
causation ID and HTTP status hoisted into columns, the duplicate
legacy Hexis trace is gone, truncated lists say so, and a rejected
credential gets its own reply instead of looking like an outage.
Found in review of #83 and #84.
setEcosystemHeaders minted a fresh correlation ID whenever the context
carried none. Every hop of one action therefore got a different ID, so
a trace could not be followed from resolve to attention to execute.
The header layer now only reads what the caller assigned. Praxis
requests are typed the same way Nexus ones already were, so a 401 from
Praxis reports as unauthorized instead of a generic failure, and a
"resolved" response with no entity is an error rather than a silent
empty result. Hexis refuses to wire at all when a token is configured,
because the vendored client cannot send one and starting anyway would
send unauthenticated calls under the belief they were authenticated.
Found in review of #84.
Traces were written as facts. A single Praxis action wrote several of
them, so machine-rate rows crowded out the bounded fact readers that
humans and evaluation consume. The habit profile window of 2000 facts
and the memeval snapshot both filled with call records instead of what
Maven learned about the owner. Traces now go to ecosystem_traces, with
correlation, causation, duration and HTTP status as columns, pruned to
the most recent 5000. The new reader is exposed over IPC and rendered
as the Calls card on the ecosystem page, so it is a table someone
actually looks at.
Found in review of #84.
The package comment, the embedder log and the startup line all said
enrolment was live and only recognition was blocked. Enroll embeds every
sample before it stores anything, so with no model on the box it fails
on the first sample with ErrDisabled and nothing is ever stored. List
then returns an empty list forever and Forget has nothing to delete. The
shipped state was three methods, all no-ops, announced as a working
half.
SpeakerConfig.Recognizes was written as the gate for this and never
called, so a block with enabled and no model_path wired everything and
skipped the one warning the operator needed. It is the gate now, and
that config shape logs why it stayed off.
Three smaller repairs. ErrDisabled had no case in speakerErr and reached
the surface as an opaque core failure, when it means the same thing
ErrUnknownMethod does. Forget read the row first and answered ErrNotFound
on a second call, so the layer documented as the one that must always
work reintroduced a failure for a voiceprint that was already gone.
And a row with unparsable metadata listed as a plausible profile named
after its own id with 0 samples, which is what a real minimal enrolment
looks like; it is reported as damaged now.
Found in review of #74.
Speaker profiles share the vector table with notes and facts. The doc
comment said reading them through Catalog is what keeps recall from
ranking a voiceprint. It is not. Catalog controls how speaker code reads
its own rows and says nothing about Search, which scanned every row.
What actually hid them was cosine returning 0 on a width mismatch, so a
192-dim ECAPA row scored 0 against a 384-dim query. Some x-vector
exports are 384-dim, and one of those would have surfaced speaker:kami
as a recall hit carrying the name of a person.
Both backends now skip the prefix in Search, and the prefix is one
constant in internal/memory so the store layer can filter on it without
importing internal/speaker.
Two more differences between the backends closed here. ByPrefix on the
in-memory store returned the stored metadata map by reference, so a
caller editing a returned Record edited the row, while the persistent
one unmarshals fresh. And the append to upsert change in Insert is a fix
in its own right, not only a speaker concern: any re-indexed id used to
leave a second stale copy searchable.
Found in review of #74.
An allowlist row stores cmd ["mcp", server, tool]. That is a late-bound
reference to a name the far end owns, so the row pins nothing about
behaviour: a server could redefine an enabled read-only list_tasks into
something that writes, and Maven would keep calling it with no confirm
turn and no second approval. Discovery now stores a fingerprint of the
declared shape, name, description, input schema and readOnlyHint, and
compares it on every refresh. A mismatch drops the row back to proposed
and, if it stopped claiming read-only, marks it destructive. destructive
is only ever raised. A row predating the column adopts its fingerprint
silently, because an upgrade is not a redefinition.
Nothing retracted a proposal either, so a tool a connected server no
longer offers stayed enabled and failed at call time with an internal
string. Those rows are withdrawn, with provenance saying why, and only
for servers that are actually connected so a restart does not disarm
what he approved.
Argument binding rested on readOnlyHint, which the same server writes.
A server advertising delete_project as read-only got an unconfirmed
argument-carrying call. Binding now also requires the tool be named in
allow_tools, something local, and refuses a required property the schema
never describes rather than guessing it is a string.
wireMCP dialled synchronously from run, and on the passkey path from
inside the unlock handler, so one black-holed endpoint delayed boot and
the answer to an unlock. The first dial happens on the refresh goroutine
under the daemon context. Two servers whose names flatten to one local
allowlist name no longer share a row.
Found in review of #71.
kami
claude