Reported: the speakers page showed 51 needing review in Chrome on Windows
and 20 in Safari on the Mac, and showed no names starting with D including
"Deal Hunter" - despite several reloads.
Neither browser was wrong. The active project is a per-browser cookie, so
each browser was on a different book and the two lists are unrelated:
Last Dwarf 91 speakers, 51 need checking, 0 D-names
DealHunterShip_4.0e 61 speakers, 19 need checking, includes
Deal Hunter, Dealer, Distorted Voice,
Duchess Victoria, Duty Nurse
"Deal Hunter" is the ship AI in the Deal Hunter book and is not a Last
Dwarf character, so its absence there was correct. The page gave no hint
that the book was even part of what was being displayed.
- The speakers page now shows a "Reviewing <project>" banner next to the
counts, so the number and the book it belongs to are read together.
- Added a project switcher directly on that page; previously changing
project meant navigating via the dashboard.
- activate_project accepts ?next=, restricted to same-site relative paths
so the switcher can stay put. Verified //evil.com and https://evil.com
both fall back to "/".
- list_projects registered as a template global, following the existing
active_project pattern; it degrades to [] on any error.
routes.py stays at its 76 pre-existing ruff errors. Suite: 684 passed, 8
pre-existing.
Follow-up to the badge fix: approving a speaker removed it from the
needs-check panel with no confirmation, which reads exactly like the
speaker being deleted. Reported as "the speaker advisor isn't shown now".
- A person-approved speaker now renders "✓ reviewed" with a tooltip saying
so. It previously rendered "✓ clear" whose tooltip claimed
"auto-accepted" - a false claim about a decision a machine never made.
- Approving confirms itself with a toast, drops the entry from the
worklist, and updates the header counts, all without a reload.
Separately, a pre-existing flake in tests/test_auth.py turned out to be a
real defect rather than test noise:
_session_secret() read the key file with .strip(), but bytes.strip()
removes ASCII whitespace from both ends. A secret beginning or ending
with one of those six bytes (4.35% of generated keys, measured) came back
one byte shorter on every read after the write - a different HMAC key,
silently, so a token signed before the flip could never be verified again.
Now read verbatim, with a length floor to reject a corrupt file.
That test also lacked the isolated_store fixture, so it read and rewrote the
real .umbra/session_secret - running the suite mutated production state.
This install's key is 32 bytes with no whitespace at either end, so no
existing session is invalidated by the fix.
Also verified: Advisor remains at display position 1 of 91 with its
utterance (utt 3668, ch15) intact - only the warning badge changed.
auth.py test flake 4/25 runs -> 0/30. Suite: 684 passed, 8 pre-existing.
Reviewer report: pressing Approve left the "needs check" badge in place and
appeared to navigate to a new page.
Root cause was a split brain, not a single bug. Approve writes the
voice_status.json ledger, but speaker_confidence read the voice_status key
embedded in speakers.json. That inline copy is stale - the web layer
overwrites it from the ledger on read (_load_speakers), so the page showed
the right value while the assessment computed from the old one. In Last
Dwarf, "Advisor" was 'pending' in speakers.json and 'approved' in the
ledger, so the reviewer's click had saved and still left it flagged.
- assess_speakers() takes the ledger as an explicit argument, authoritative
over the inline key; load_speaker_assessments() reads voice_status.json.
- New SpeakerAssessment.reviewed flag, surfaced in the serialized view.
- Approval clears the badge but keeps the reasons visible: the underlying
facts (role-only name, one utterance) stay inspectable on hover rather
than being silently discarded.
- Corrupt/ non-dict ledgers degrade to {} - this module never raises.
- Approve swapped #speakers-list with innerHTML, replacing the whole
reviewer list with a single <tr>; now swaps only its own row via
outerHTML on #speaker-row-N.
- The review panel explained only rename/merge and never mentioned Approve,
so the process was unclear; it now lists the three outcomes.
16 new tests. Suite: 681 passed, 8 pre-existing failures unchanged.
Reviewing speaker exceptions meant clicking Regenerate, watching a modal that
blocked the page, then navigating back to find out what happened. Worse, the
route refused a second request with a 409 while one was running, and
/api/generation/status exposed only a single global `active` flag - so there
was nowhere to put a second speaker and nothing to render per row.
Now: POST returns as soon as the task is captured, the row shows a status tag
(queued / working / done / failed), and further speakers can be queued while one
generates. No navigation, no modal.
umbraaudio/voices/voice_queue.py
- Ordered, serial queue. TTS is strictly serial here, so items run one at a
time in enqueue order; the second speaker waits rather than contending.
- Per-item state, persisted to voice_queue.json. An item left `running` by a
crash is marked `error` on load - the honest outcome, since silently
re-running could write audio nobody asked for twice.
- Enqueueing a speaker that is already pending is a no-op returning the
existing state, so a double-click cannot queue the same work twice.
- `cancel` withdraws a queued item but refuses one already running.
- Status vocabulary (queued/running/done/error) is deliberately small and
stable: it is what the controller will map onto node task states.
umbraaudio/web/voice_queue_routes.py - POST /api/voices/queue (capture the task),
GET /api/voices/queue (report per-item state), POST .../{speaker}/cancel. Kept
out of the hot routes.py, following name_token_routes.
/api/speakers/regenerate-voice is now a thin shim onto the queue so a
bookmarked call still works and the two paths cannot race. Net -24 lines in
routes.py; the blocking implementation is gone, not commented out.
The UI's modal, its 409 handler and the page-load resume handshake were removed
(67 lines) rather than left dormant - queue state is re-read from the snapshot
on load and after each HTMX row swap, so it survives navigation.
Unknown speaker names are rejected with a 404 at enqueue time. Previously they
reached the profile builder and surfaced as a bare KeyError('primary_name') in
the failed tag.
Verified live: two speakers enqueued back to back, first ran while the second
stayed queued, then the second started automatically on completion; both wrote
real ref WAVs. 16 new tests. Full suite 665 passed, 8 pre-existing failures.
The UI had no authentication at all: `active_project` is an unsigned cookie
and the server binds 0.0.0.0, so every page - including the mutation endpoints
and the audio downloads - was reachable by anything on the LAN and, through
the Caddy reverse proxy, from the internet. Spec 21/22 recorded this as an
unauthenticated mutation surface; this closes it locally until the controller
on oldbox takes over identity.
Gate is middleware, not per-route dependencies, so a newly added route is
protected by default and forgetting a decorator cannot expose the UI. It is
fail-CLOSED: an exception while checking a session denies the request, because
falling through to "no session required" would turn a bug into an open door.
- umbraaudio/auth.py: PBKDF2-HMAC-SHA256 (stdlib, per the project's
stdlib-before-third-party rule) with a random per-user salt. Sessions are
HMAC-signed and carry no secret, just username and expiry. All comparisons
use hmac.compare_digest. Credential writes are atomic (tmp + os.replace) so
an interrupted save cannot lock the user out, and both the store and the
session secret are chmod 0600 under .umbra/, which is gitignored.
- First use: seeds username `bdumbra` with the one-time password `First321`,
stored hashed and marked must_change, so the bootstrap password never lands
on disk in plaintext and is dead after the first login.
- Password rule: more than six characters, at least one letter and one special.
- Sessions last 14 days, sliding.
- Login failures are deliberately vague: a wrong username and a wrong
password produce the same message, so the form cannot be used to enumerate
accounts.
Verified live: unauthenticated requests to /, /speakers, /name-tokens, the
export endpoints and the ZIP download all 303 to /login; the first-use password
forces /change-password before any other page; weak, mismatched and reused
passwords are rejected; the old password stops working; tampered, forged and
expired session tokens are all rejected. Static assets remain reachable so
the login page renders. 19 new tests. Full suite: 649 passed, 8 pre-existing
failures.
The dashboard had per-track MP3 links but no way to get the package as one
artifact, which is what a distributor actually wants. New
GET /api/export/download-package.zip streams every track plus ar_package.json
so the download describes itself: order, roles, durations, measured RMS/peak.
Stored, not deflated - the MP3s are already compressed, so deflating 614 MB
would burn CPU for no gain.
Two bugs found while verifying rather than assuming:
- buffer.seek(0) sat INSIDE the ZipFile context manager, so it rewound before
the central directory was written. Every download came back 1,763 bytes
short and unzip could not find the end-of-central-directory record. Moved
the seek after the block closes; now "No errors detected" and all 32 MP3s
plus the manifest compare byte-for-byte identical to the originals.
- strict= is a zip() builtin parameter, not a ZipFile one. Ruff's B905 nudge
had me add it to ZipFile(), which raised TypeError and produced an empty
200. Reverted; the one remaining B905 is pre-existing elsewhere in the file.
Also surfaced the accepted_deviations from the manifest in the dashboard: a
collapsible table of the tracks below the RMS floor, with their measurement,
peak, crest factor and the rule they depart from, so the exception is visible
where the numbers are rather than only in the JSON.
Verified error paths: no active project -> 400, project without a built
package -> 404. Lint unchanged (77 before and after). Full suite: 630 passed,
8 pre-existing failures.
Single master MP3 set for both distributors, with the two sub-floor chapters
left as they are rather than limited.
accepted_deviations in the manifest, for tracks that are peak-compliant but
still under the RMS floor. Both sit at exactly -3.00 dB peak, so there is no
headroom to add the ~3 dB of gain they would need: a 20-21 dB crest factor
makes RMS -20.5 and peak -3.0 mutually exclusive at this gain. A limiter would
meet the number by squashing transients, which is a worse trade for a
narration-led book. Recorded as a decision with its measurement and reasoning
so a reviewer reads it instead of rediscovering it from a warning string.
Deal Hunter: "The Poker Game Distraction" (-23.80, crest 20.80) and
"Hunting Pirates" (-23.10, crest 20.10). The 30 compliant tracks span
crest 17.60-19.90, so the two separate cleanly and nothing marginal is swept
into the exception.
Also quarantined three orphan WAVs in ch35 (utt_0003/0004/0005, 40.9 s, dated
2026-08-29). They predate the closing track being rewritten to three
utterances, have no marker and no DB row, but the stitcher globs utt_*.wav -
so they were silently included in ch35_full.wav and in the shipped m4b, whose
closing chapter was 64.5 s where the tracked audio is 20.3 s. Moved to
chapter_audio/_orphans_ch35_20261009/ rather than deleted. Only ch35 was
affected; the book-wide orphan scan found no others.
Rebuilt both artifacts from the cleaned chapters: 32 MP3s all peak/rate/bitrate
compliant, and the m4b closing chapter is now 21.3 s.
_encode_and_enforce_peak measured RMS once, before any peak-correction trim,
then returned that stale figure. A track that needed peak correction was
therefore recorded louder than it actually is, and loudness_ok passed on a
number that described neither the intent nor the artifact.
ch15 took a -2.9 dB trim: the manifest said -20.6 dB, the delivered file
measured -23.8 dB, and the gate called it compliant. Measuring all 32 delivered
files independently found 2 real failures (ch15 -23.8, ch21 -23.1, both under
the -23 dB floor) that the manifest reported as 0.
Now both figures come from the encoded file after every trim, so the manifest
always describes the delivered bytes.
_measure_file gained a content_s argument that excludes head/tail room tone
from the RMS measurement via atrim. RMS is defined over "the entire program
(not including room tone)", and padding share varies enormously: 0.03 dB on a
30-minute chapter but ~7 dB on a 1-second fixture. Peak needs no such
treatment since padding is silence and cannot raise a peak, so one trimmed
pass serves both.
Rebuilt: 32 tracks, 7.45 h, 0 peak failures, loudness range -23.80..-20.70 -
which now matches independent measurement exactly.
The delivered m4b was 24,348s where the source chapters total 26,830s. Root
cause: libsndfile stops reading at 2**30 - 1 frames, and a 7.4-hour book at
44.1 kHz is ~1.18e9. ANY read through libsndfile was capped - including a
chunked read loop, because the limit is in the library's frame counter rather
than in the call. ffprobe read the same file correctly at 26,830s, which is
what made it look like the concat had lost data rather than the read.
Both ends now stream through ffmpeg instead:
- _concat_wavs_with_silence writes with sf.SoundFile per segment rather than
one np.concatenate() array, and as FLOAT rather than the PCM_16 default,
which had been clipping the book to 0.00 dBFS before normalisation saw it.
- _encode_aac_m4b measures RMS with ffmpeg and encodes with a
volume+alimiter filter chain, so libsndfile never reads the book at all.
alimiter looks ahead, so it holds the ceiling without the flat-topped
distortion hard clipping produces, and it survives the AAC encoder better.
The peak margin also moved from a post-hoc trim to the limiter ceiling. Trimming
after the fact cost ~3.2 dB and pushed RMS to -23.8, outside the band; clipping
to -6.0 dB instead lands at -3.10 peak / -20.70 RMS with no trim at all.
Verified on the real book: duration 26,799s (7.44 h), AAC 44.1 kHz mono
125 kbps, RMS -21.1, peak -5.2. 32 chapter markers, cover art embedded,
closing credit last. Full suite: 630 passed, 8 pre-existing failures.
export_m4b encoded with `-c:a aac -b:a 128k` and no `-ar`, so the delivered
file came out at 24 kHz / ~96 kbps mono while the AR MP3 masters were
44.1 kHz / 192 kbps CBR mono. Measured on the shipped Deal_Hunter.m4b:
24 kHz, ~96 kbps, and peak -2.0 dB - failing the 44.1 kHz and -3 dB
requirements at Author's Republic and Spotify. AudiobookExporter had zero
loudness or peak handling of any kind.
- _encode_aac_m4b now passes -ar 44100 and -ac 1 explicitly, and takes the
spec constants from ar_export so the two paths cannot drift apart again.
- Added normalisation to the RMS target and peak enforcement on the delivered
file, plus m4b_loudness_report() so the result can be verified.
- AAC overshoot is NOT monotonic. Measured +0.8, +0.9 then +1.6 dB across
three passes, so trimming by the exact excess oscillated and exhausted the
pass budget still breaching. Trimming by excess + 1.5 dB margin converges:
verified -2.20 -> -4.5 dB in two passes.
- _concat_wavs_with_silence now targets 44.1 kHz rather than the TTS-native
24 kHz.
- Closed spec 22 gap 1: _stitch_chapter hard-coded
silence_between_chapters_ms=0 while ChapterBuilder's AudioStitcher defaults
are 500/4000, so the packaged book had different silence than the audio a
reviewer approved.
Trade-off, flagged rather than hidden: the 1.5 dB overshoot margin costs ~1.3
dB of RMS, so a trimmed m4b measures about -22.8 dB against the -20.5 target.
Still inside the -23..-18 band, but only ~0.2 dB from the floor; a chapter
needing more than ~2.5 dB of trim would fall out of band and loudness_ok will
flag it.
Verified on real chapter audio: AAC 44100 Hz mono, RMS -22.8, peak -4.5.
6 new tests. Full suite: 630 passed, 8 pre-existing failures unchanged.
The compliance gate certified our intention rather than the artifact.
_encode_track measured the pre-encode float buffer, and _normalize
hard-clipped to exactly -3.0 dB, so all 32 shipped tracks recorded
peak_db = -3.0 and peak_ok = True. libmp3lame then overshot the ceiling, so
the MP3s a listener actually received peaked near -2.0 dB - above the -3 dB
limit that Author's Republic, INaudio and Spotify all require.
- _measure_file() measures an encoded file; the gate now reads the MP3.
- _encode_and_enforce_peak() re-encodes with the excess trimmed until the
FILE complies, so codec overshoot is corrected rather than assumed away.
Measured on real audio: -2.0 dB before, -3.1 dB after, and volumedetect,
ebur128 true peak and the gate all agree.
- RMS is still taken from the content before padding, because every
distributor defines RMS "across the entire program (not including room
tone)". Measuring the padded file read ~9 dB low (-29.4 dB on a title that
normalizes to -20.5). Peak comes from the encoded file, since silence
padding cannot raise a peak and only the file reveals the overshoot.
- Same fix applied to encode_chapter_mp3, which shared the defect.
- Closing credit is now emitted LAST. The roles loop emitted
("sample","opening","closing") before the chapters, so the end-of-book
credit sat at position 3, ahead of the story it was meant to conclude.
Tests: added coverage asserting the recorded peak equals an independently
measured peak of the delivered file, that no file exceeds the ceiling, and
that a clipping source is trimmed until it complies. Three existing
assertions had encoded the old behaviour and were corrected - including two
"-2.9" tolerances that existed only to accommodate the overshoot, and an
ordering assertion that pinned the closing track second. Those now assert the
real -3.0 limit and the correct order.
Full suite: 624 passed, 8 pre-existing failures unchanged.
The box is offline and being reconfigured. Removed the relay helper URL from
.env (qwen_tts_helper_url defaults to empty, so local :8881 is now the only
backend), dropped the 18881 relay entry from MEANINGFUL_PORTS, and made the
node report state plainly that no controller endpoint exists rather than
advertising one on a retired host.
Also recorded that ModelCoordinator.ensure_qwen_server() already spawns the
Qwen server on demand - it checks /health first, so a launchd-owned instance is
reused instead of double-spawned.
launchd scans $HOME/Library/LaunchAgents once at login, and $HOME is the
relocated /Volumes/Dock/worstauthor. On 2026-10-08 the Dock volume mounted at
05:59 but login was 05:58, so the plist was invisible and com.umbraaudio.web
never loaded - the web server stayed down after a reboot.
Installed to /Library/LaunchAgents instead (system-wide, on the boot volume,
scanned regardless of $HOME). Removed the $HOME mirror since two plists with
one Label is a race; backup kept. Verified running from the new path.
The banner already warned that voice generation "will fall back to Edge-TTS",
but a warning is not a lock: voice_maintenance was wired into routes.py purely
as a template global, so nothing actually stopped generation. Clicking Generate
during an LLM attribution run would retry Qwen, exhaust the attempts, write a
degraded Edge-TTS ref to disk, and record it in voice_engines.json. For a
quality-first pipeline that is worse than refusing.
_add _voice_maintenance_block(), returning 409 with an explanation of why
(Qwen3-TTS cannot co-reside with the LLM on this machine), and wire it into the
six endpoints that synthesize audio: generate_utterance_audio, delete_utterance,
create_utterance, regenerate_speaker_voice, generate_all_speaker_voices,
generate_full_audiobook.
Deliberately NOT blocked: speakers_list_partial (a Qwen health check, i.e. the
banner the reviewer needs) and set_voice_status (reads an existing _ref.wav to
capture a vmatrix, part of the approval workflow). Blocking those would have
prevented auditing and approving voices while analysis is paused.
The marker is owned by a pid and checked for liveness, so a crashed run
self-heals and cannot strand generation permanently.
Records the access mechanism (osxkeychain credential helper + Keychain-stored PAT,
deliberately not the token value), the private/login-required repo posture, the
push order, the shallow-clone fix before mirroring Qwen3-TTS, the :2222-vs-:22
Gitea SSH trap with its separate host key, the forced-password-change gate that
hid the SSH key settings, the mini-eq .18/.138 addressing quirk, and the LAN
dependency the submodule repoint introduces.
Two groups of changes, committed together as the follow-up to the initial push:
1. .gitmodules — submodule Qwen3-TTS now points at the internal Gitea fork
(ssh://git@mini-eq:2222/umbragit/Qwen3-TTS.git) instead of upstream GitHub,
so a clone actually gets the UmbraAudio branding/integration commit
012e452 rather than vanilla upstream.
2. New infra/ tooling for the cross-machine work in spec 21: agent_bridge,
remote_credential_broker (Keychain-backed, secrets read at runtime and
generated via read_secret/random_secret — no hardcoded credentials), the
enterprise bootstrap scripts, and the oldbox SSH/auth prompts. Plus
backup/ restore scripts, two specs (18 UmbraTrans, 19 enterprise control
plane), umbraaudio/voices/lmstudio_gate.py, and in-flight edits to the
discovery/attribution/profiling modules, config, chapter_builder, emotion,
pronunciation, qwen_tts and routes.
Also adds *.db to .gitignore: SQLite project databases are runtime state and no
.db file was tracked before.
Verified before staging: .env's only change is UMBRA_TTS__QWEN_TTS_HELPER_URL and
the committed copy has zero secret-looking values; the broker ships only
config.example.json.
Supersedes the long-standing 'SSH denied, needs manual provisioning' note, which
was wrong in its diagnosis. Recorded: root cause was authorization (our ed25519
public key missing from umbra's authorized_keys on oldbox), not networking,
TCC, or host keys; the local key + fingerprint; all three pinned host-key
fingerprints verified matching; the Host oldbox config block and its backup; the
remote-broker profile now returning exit 0 (and the oldbox-umbra vs
oldbox-umbra-ssh naming trap); the read-only fingerprint-verification recipe and
the rule not to pin a handed-over fingerprint unchecked; and oldbox's specs
including the NVML permission gap that blocks nvidia-smi for umbra.
The operator's chapter 6 review found a line attributed to Hallr that the text
gives to a scum boy, and no part of the system had flagged it. Spec 24 §2
diagnosed why: the duplicate scan only compares names that already exist, so it
cannot see a line given to the wrong EXISTING speaker. Spec 24 §4A proposed a
mechanical evidence gate. This is that gate, measured.
New umbraaudio/analysis/speech_evidence.py, model-free, plus
scripts/report_speech_evidence.py (read-only) and 44 tests.
For every dialogue line: extract the speaker the narration actually names, build
the candidate set from the cast rather than from the guess, and return
supported / contradicted / unsupported / unconstrained.
Last Dwarf, all 49 chapters, 3,918 dialogue lines:
supported 1,844 47.06%
unconstrained 2,032 51.86%
contradicted 34 0.87%
unsupported 8 0.20%
review queue 42 1.07%
It catches the chapter 6 defect exactly: "The leader on the right for Hallr
says," / "We want to be the ones to claim we killed the last dwarf" assigned to
Hallr, with the two leaders who could own it. attribution_qa reports 11 findings
book-wide and this one is not among them.
Two design decisions carried the precision, and both were forced by measurement.
First, "no tag" is not a finding: most dialogue has no tag because a tag
introduces only the first line of a conversation, so unconstrained and
unsupported are separate verdicts. The first working version queued 60% of the
book for this reason, which is the same as having no gate. Second, "ambiguous"
was removed as a verdict - it hid the case that matters, since a line assigned to
Hallr under "The leader on the right" is wrong no matter how many leaders exist.
Candidate count now decides whether a suggested fix is confident or a question.
Two corrections to spec 24, both from measurement. The 20 role-tag violations
reported in section 3 are really 11; my ad-hoc probe was worse than the checker
it was meant to replace. And identity coverage is already 100% - every dialogue
line resolves to a profile or alias - so the defect is never a missing speaker,
it is a line given to the wrong existing one.
Six extractor bugs the measurement exposed are now regression tests: a role
phrase with a spatial qualifier produced no tag at all; a coordinated predicate
spoke as its first subject instead of its last; a possessive voice resolved to a
profile called Voice; irregular plurals and superlatives never matched; and the
subject was read off trailing participial clauses. Two generalisable rules came
out of them - a speech tag's subject is a noun phrase and a prepositional
qualifier is still part of it, and a subordinating conjunction opens a clause of
its own ("The dragon swallows when Hallr says" is Hallr).
Reports carry chapter_id AND title because chapter_id and chapter_index differ
by one in every project here; a bare "ch8 utt264" sent the operator to the wrong
chapter, and the line they asked about is chapter 9 utt264.
Full suite 619 passed / 8 failed, the same 8 pre-existing failures. ruff clean.
The operator reviewed ch6 with author knowledge and found real defects the
duplicate scan could not see. Two corrections and one structural diagnosis.
I was wrong about two pairs. I wrote "probably one person" about The Officer /
Officer, and "probably two different riders" about Border Soldier / Soldier
(Rider). Both are two different characters. Measured, The Officer (113.3 Hz) and
Officer (104.5 Hz) are 1.4 SEMITONES apart - two distinct characters with
near-identical voices, which is a genuine cast-clarity risk that a merge would
have hidden. Border Soldier and Soldier (Rider) are 12.2 semitones apart, so
their risk is narrative proximity, not vocal.
The real lesson is the principle behind it: because the author wrote the book he
knows these facts, and on a book he did not write nobody does. So every merge
decision must be resolvable from evidence in the text, and the default when
evidence is absent must be SEPARATION. Keeping two speakers separate costs one
redundant voice; merging two people costs a wrong character for the rest of the
book. Reworded the scan so a blocked group is reported as "same name-space,
different characters unproven" and never as a probable merge.
Why the scan missed ch6: it can only compare names that already exist. It cannot
see a line attributed to the wrong EXISTING speaker. ch6 utt29 - "The leader on
the right for Hallr says," / "We want to be the ones to claim we killed the last
dwarf" - is a scum boy attributed to Hallr. No duplicate existed to find. The
scum group was also missed on identity: only the case variants were caught, and
nothing compares a designation against the text that produced it.
Diagnosis, measured book-wide: a role speech-tag must not resolve to a distinct
named character. 20 dialogue utterances violate it. True positives include ch11
utt191-193 where ONE tag ("The young man replies,") was split across Kaleb,
Tobias and Hallr, and ch13 utt45 where "The girl replies," became The King. But
40% are false positives because ROLE_WORDS is incomplete - royal, king, witch,
doctor and others are missing - and a gate that noisy would be ignored. The
vocabulary must come from definitions.py, not a hand-written list.
Plan, model-independent first: (A) mechanical evidence-coverage gate, which
catches this class with certainty; (B) constrained assignment - choose from the
evidence-supported candidates only, not all 93 names, which is the actual cause
of utt29; (C) role-phrase preservation so "the leader ON THE RIGHT" stops
acquiring a new name per mention; (D) tag-run consistency, never split one
introduction across three characters; (E) a segmentation gate for narration typed
as dialogue (ch6 utt26); (F) confusability signals, a casting concern distinct
from identity; (G) an evaluation harness with a ch6 gold set and five metrics,
because without a number "use a better model" is a guess.
On a better model, honestly: 12b cannot fit in 16 GB, and the GPU box at
192.168.123.14:1234 refused a connection today (curl exit 7), so it is blocked.
And B removes the search space first, which is where the error actually was.
Full suite unchanged at 575 passed / 8 failed, the same 8 pre-existing failures.
The operator confirmed Narratorold and OrigNarrator are dead profiles. --force
exists for exactly this: a decision the scan is not entitled to make. It prints
the overridden conflict before acting, and never deletes a reference file.
Applied to Last Dwarf:
profiles 93 -> 91
approved voices 13 -> 13 (unchanged)
voice ref files 72 -> 72 (unchanged, both WAVs left on disk)
dialogue rows labelled 'narrator' 3 -> 0 (now the approved Narrator)
narration rows labelled 'narrator' 5978 -> 5978 (untouched)
The dead names survive as aliases of Narrator so no stray utterance can
resurrect a profile, and the 3 re-pointed rows are a genuine attribution fix:
ch5 utt26 and ch27 utt125 are narration mis-segmented as dialogue; ch42 utt290
is a first-person line that may be intended inner thought.
Rescan: 10 auto, 2 review, 3 blocked, unresolved spoken names 3 -> 2.
Full suite 575 passed / 8 failed, the same 8 pre-existing failures. ruff clean.
Last Dwarf has accumulated duplicate speakers from successive discovery and
designation scans: The Girl / The Girl helping / One Girl Helping, Guard 1 / The
first guard, Walking oldest / The Oldest, plus three lowercase spoken names that
differ from their profiles by one letter. The operator wants them eliminated and
was explicit that created and approved voices must not be lost.
umbraaudio/analysis/speaker_dedup.py
Read-only fact gathering (profiles, ledger, voice refs, chapter spread) and a
matcher that explains itself: determiner and ordinal normalization, superlative
stemming, and three VETO rules, because string similarity alone would merge
people who are not the same person:
- distinguishing modifiers: The Oldest is not The Tallest; Guard 1 is not Guard 2
- role nouns: Soldier (Rider) is not One Rider
- proper names need ratio >= 0.88: Anfisa (0.77 from Tanisha) is a different
character; Rielan/Rieland at 0.93 is a misspelling
Verdicts auto / review / blocked. A group with two different voice references is
blocked, because merging would discard a generated voice. Survivor selection
prefers an approval, then a voice reference, then utterance volume.
Two rules were written and then removed, recorded in the spec so nobody re-adds
them: a "shared token plus role noun" rule proposed The Oldest = The Old Woman,
and a "single shared token" rule proposed Scum leader = The Leader. A missed
duplicate costs one manual merge; a wrong one silently rewrites a speaker.
Two real bugs, both found by running against a COPY of Last Dwarf rather than by
reading code:
1. orphan_resolution.merge_speakers never touches voice_status.json, so merging
an approved speaker into an unapproved target would have deleted the approval
record silently. merge_group() now carries the approval to the survivor,
drops the loser's entry, and reports approvals_carried.
2. The first re-point was unscoped and moved 5,981 rows instead of 3: `narrator`
is the labelling convention for ~5,978 NARRATION rows too, and renaming them
would invalidate thousands of clips because every marker records the speaker
exactly. The re-point is now dialogue-only. Narration casing is a separate,
deliberate decision and is flagged as such, not taken by a duplicate scan.
Verified on a copy: 93 -> 84 profiles, 23 utterances re-pointed, narration
untouched, 72 -> 72 voice ref files with zero removed, 13 -> 13 approvals with
zero lost.
Last Dwarf scan: 10 auto, 2 review, 4 blocked. Blocked groups are exactly the
ones the operator must decide - The Girl (two voices for one person), Officer,
the three-way Narrator family, and Border Soldier vs Soldier (Rider).
Nothing was applied to the live project. tests/test_speaker_dedup.py: 37 tests,
vetoes weighted as heavily as positives. Full suite 575 passed / 8 failed, the
same 8 pre-existing failures. ruff clean.
Two corrections from the operator, and I had both wrong.
1. "I didn't explicitly approve all minor character voices. By releasing the
audiobook I have given implicit approval those voices are ready and should be
tagged." The previous session reported "24 of 61 speakers have an approval
record; 38 do not" as if it were a gap. It is not a gap. Releasing IS the
approval, and those voices were waiting to be tagged.
chapter_approved and audiobook_approved now read true, with
approved_by: "this publication record" - there is no separate audiobook
approval in this system, and pretending otherwise understated the release.
The boundary matters, so it is measured rather than assumed: implicit
approval covers voices that HAVE LINES in the released chapters, computed from
the database. A character who never speaks is not covered, because releasing a
book does not approve a voice it never uses. On Deal Hunter edition 1 that is
24 explicitly approved, 29 implicitly approved, and 10 in-profile voices with
no lines. The three categories are never summed - an explicitly approved voice
can also be one that stopped speaking.
grant_implicit_voice_approval() writes plain status strings, the shape the web
UI already reads, and keeps the provenance in the immutable publication record
rather than the mutable ledger, so it cannot be edited away.
2. The m4b inter-chapter gap should be the same 2 s as the lead and trail
padding. export.py had gap_ms=4000 in one path and a default of 1000 in the
other, while ar_export.SILENCE_MS = 2000 governed lead and trail: three
drifting values for one concept. The gap now imports SILENCE_MS as
CHAPTER_GAP_MS, so all three are one number that changes in one place.
New umbraaudio/release_comments.py - the comment is a gate, not a caption.
designator 2026-10-03 #2 14:32:05
slug release-20261003-143205-02
The sequence is per date, not global, which is the whole point: a problem found
and corrected on the same day is unambiguous about which was last. Two comments
authored in the same second still get distinct slugs, because the sequence is
allocated by scanning the directory. First line is the summary, the rest is the
body, like a commit message. An approved comment is FROZEN - changing an approved
note is a different claim about a different state, so it needs a new comment and
therefore a new publication.
require_approved() blocks --publish: passing a draft exits 2 and says what to do
next. Verified live.
draft_comment_text() composes the comment from the measured build rather than
from memory. Deal Hunter's draft is written and waiting:
release-20261003-133508-01, 32 chapters, the 24/29/10 voice split, 32/32
loudness and peak gates green, and the post-approval padding note.
New invariants R-8 through R-11 in spec 22. Two of my own test premises were
wrong and the implementation was right - recorded in AGENTS.md because the tests
were the weaker artefact.
Full suite 538 passed / 8 failed, the same 8 pre-existing failures. ruff clean.
Nothing was published: the Deal Hunter comment is a draft awaiting review, and
voice_status.json is untouched because publishing never ran.
The operator clarified the release model: the system keeps current and previous
state, there must be an explicit publish flag the user sets, there are approvals
for a speaker voice but none for the audiobook or chapters, the chapters that
make the m4b are the versions to flag, and a fix found after assembly causes a
change rather than editing what shipped.
Deal Hunter is the precedent, and it is measurable rather than theoretical.
Against the real package (built_at 2026-09-09T11:21:11Z), chapter 6:
ch6_full.wav 2227.26 s @ 24000 Hz what a reviewer listens to
Deal_Hunter_01_Salvage.mp3 2231.26 s @ 44100 Hz what a listener received
+4.00 s = 2 s lead + 2 s trail
plus a resample and a -20.5 dB RMS normalization. Found on 32 of 32 chapters.
The approved artifact and the published artifact were different files, and
nothing in the system recorded either fact. ar_package.json is rewritten by
every rebuild, so it is a build status, not a release record.
umbraaudio/release.py
DeliveryProfile.from_ar_package reads the profile the exporter actually used,
so the record follows the code instead of a declaration that can rot.
build_publication is read-only and pins every chapter wav, every mp3 and the
package by full SHA-256, plus a per-chapter source_unit_key_digest so a release
pins which source revision produced it, not merely which bytes.
verify_publication returns ok / drift / warnings and never silently accepts a
changed artifact. write_publication is atomic and refuses to overwrite an
existing publication id - immutability enforced by the writer, not by
convention. A later fix is edition n+1 with supersedes pointing at n.
The rule that prevents a repeat: you cannot approve a chapter and then pad it.
An approval must name the artifact a listener receives. Every record states
chapter_approved: false and audiobook_approved: false in its own words, because
that is what would have caught Deal Hunter.
The record also refuses to overstate approval. voice_status.json is an approval
LEDGER - it only holds speakers someone explicitly touched - so it has 24 entries
against 61 speakers in the profile. Reporting "24 approved" as if it meant "all
approved" would claim an approval nobody gave, so the counts are separate and the
38-speaker gap is named.
scripts/publish_release.py: dry run by default; --publish, --verify, --list,
--edition/--supersedes. Exits 2 while blockers exist.
Also found, reported and NOT fixed (specs/22 section 7 gap 1): export_m4b's
_stitch_chapter hard-codes 500 ms/0 ms silence while ChapterBuilder uses
AudioStitcher defaults and AudioConfig holds its own values, so the m4b path can
stitch a chapter with different silence than the approved one. Until that is
closed an m4b release record would pin the wrong bytes, so it is better not to
offer one yet.
And projects.status is never advanced to 'released' in any of the four live
projects - all sit at parsed/analyzed - so the publication record is currently the
only release signal in the system.
tests/test_release.py: 24 tests. Full suite 512 passed / 8 failed, the same 8
pre-existing failures. ruff clean. No export, TTS or project data was modified.
The controller (oldbox) sent a data-model alignment proposal and asked for six
additive changes on this node, the validation protocol, and an honest report.
Six of the seven changes are done; the seventh is declared missing rather than faked.
Stage 0 - reachability verified exactly as documented: health 200, heartbeat 401
(the correct answer - auth is refusing), missing Host 400. Nothing else was sent.
source_unit_key (controller 5.1, the blocking item)
umbraaudio/source_units.py. sha256("uk1" | chapter_source_text_sha256 |
occurrence | text_sha256). Nothing positional in the key: "occurrence"
disambiguates repeated text in document order.
- 18,181/18,181 units keyed across all four projects, 0 collisions
- chapter 1 renumbered +100 -> 0 keys changed (the load-bearing proof)
- backfill 10,015 rows in 0.11s, second run stamps 0 (idempotent)
- alignment phase 3: 200/200 probe units resolved with used_index: false
- tolerates the older projects that have no chapters.chapter_index
Recorded at generation time in chapter_builder, BEFORE the rows are read and
BEFORE the marker is written, so every path that renders audio also produces
identity and the clip can name its source unit.
5.7 recorded: utterance_index is KEPT as a display ordinal and made
non-load-bearing. spec 21 §10.
Blob hashes and effect chain (5.2)
Markers now carry the full sha256 of the delivered bytes and the effects that
produced them. Verified the existing order is already correct - write_marker
runs AFTER apply_effects - and closed the remaining hole: a changed trait list
changed the hash with no recorded cause. sha256 verification is opt-in, since
hashing every file on every validation pass would re-read gigabytes.
Vocabularies published (5.4) - umbraaudio/definitions.py, 19 sets, 126 values,
each value carrying whether it is 'declared' (spec/controller) or 'extracted'
(read out of code), so a diff shows a difference of meaning, not of wording.
Alignment report tool (6.6) - scripts/alignment_report.py --phase all. Emits the
controller's envelope with the per-phase evidence shapes, and obeys the two
rules they set: no evidence means blocked, never pass; a field we cannot produce
is null PLUS a finding, never omitted. Phase 2 is deliberately 'fail' - the
three-way store disagreement is real, and a pass there would be a check that
passed by doing nothing.
Fingerprint discipline (5.5) - the drift detector reported its OWN fingerprint
change when source_unit_key was added rather than hiding it. Node profile
updated with the new fingerprints and an honest note that they moved.
Node report extended - per-project source-unit-key coverage, per-voice-asset
sha256 + object_key, definition_set_version, controller stage reached.
Six gaps found in the proposal (spec 21 §11), including the one they invited plus
five more. The sharpest: "hash at generation" ignores that audio is
post-processed afterwards, and our key is anchored to chapters.source_text - a
second copy of the manuscript - so any change to chapter storage silently
invalidates every key in that chapter.
5.6 deliberately NOT faked: personality has no versioned artifact here, only
unversioned LLM prose in speakers.json. Reported as phase 6b finding B-03
instead, because wrapping unversioned inputs in a version field would produce
artifacts claiming a lineage they do not have.
Answered the controller's two open questions (infra/controller-handoff-2026-10-03.md):
forward-only link for a revealed recurring character, and yes - a new edition may
contradict a frozen state, since this node's correction path IS a regeneration.
Full suite 484 passed / 8 failed - the same 8 pre-existing and unrelated failures.
ruff clean on every file touched. The parallel session's chapter_builder hunk
(helper_url) was filtered out of the commit and left intact in the working tree.
A parallel session is building an agent-comm stdio MCP server with an
append-only NAS mailbox (infra/agent_bridge/, uncommitted). Rather than
proposing a second channel, spec 21 now maps the task/result/requirement
envelopes onto its message kinds and adopts its sanitization rules as system
rules.
Also states the boundary the parallel work should not cross: a mailbox is a
transport, not a scheduler. It carries tasks; leases and heartbeats must not
become 'a file in a folder'.
A parallel effort is moving UmbraTrans into the main Umbra system on oldbox
(192.168.123.14). This commit makes this machine describable to that system and
records what the system still needs to decide. Everything is a proposal for the
controller to ratify; no architecture is decided here.
specs/21-umbra-system-nodes-tasks-and-changes.md
Four workstreams, each with what exists, what is missing, and a proposal:
- WS1 data model alignment / copy / verify / transition. Design is far along
already (manifest.md + spec 20); the missing piece is ratification.
- WS2 every machine is a node that gets assigned any task. Job types, leases
and pools exist (spec 19 §12); the node half does not. Proposes node
identity, capability advertisement, task/result envelopes, and refusal as a
valid audited outcome.
- WS3 cross-product change management. Nothing exists today. Proposes a
requirement catalog owned by the controller, four change classes, and a
distribute-then-confirm-with-evidence loop. Includes the worked example of a
controller-owned entity (designation, pronunciation, voice attribute) that
must apply across every product.
- WS4 UI topology. Recommends hybrid: controller gets the system portal,
the audio review UI stays on the worker because it is artifact-native and
has no authentication at all. Adds a git-tag deploy + verify + confirm flow.
Also: 25 questions grouped by who can answer, each naming what it blocks, and
a task list for this machine (NOW / NEXT / LATER) where every task has a
machine-checkable acceptance criterion.
infra/umbra-node.json
This machine's declared profile: node_id umbraaudio-m4, role worker, services,
capabilities, resource pools, constraints, project list with lifecycle, schema
fingerprints, and an honest facts_for_the_controller block (empty tables, no
stable utterance id, no auth, no shared-postgres client). No secrets - ids,
names, labels and counts only.
scripts/umbra_node_report.py
Read-only node report. Git state, listening ports, .env KEY NAMES only, per
project row counts, and a schema fingerprint hashed over sorted
table.column:type so ALTER TABLE reordering does not read as drift. --json for
machine consumption; --check compares declared profile to measured reality and
exits 2 on drift. Drift is deliberately limited to stable facts (contract
version, project set, schema fingerprints) - git HEAD and live ports are
reported but never drift.
manifest.md
Now explicitly a dated measurement with an update log, because other machines
read it as current. 2026-10-01 refresh: Last Dwarf speakers.json 70 -> 93
after the designation merge, DealHunter ref WAVs 94 -> 97. Every other §1.1
count re-verified unchanged.
Measured this machine: 4 projects, 2 distinct schema fingerprints (real
per-project drift), only 4 of 11 tables populated, 461 tests passing / 8 known
pre-existing failures. No UmbraAudio behaviour changed.
Ran discovery on Last Dwarf ch5 with the new over-detection prompt rules. It
improved (found 'The big one', 'The one on the left/right', 'The Leader',
'Girl helping' where it previously found none of them) but STILL dropped
'The oldest' and 'The tallest' from the same six-man crowd — the e4b model is
not deterministic enough to guarantee coverage.
So added a mechanical backstop that does not depend on the LLM at all:
- umbraaudio/analysis/designation_scan.py reads the three canonical
speech-attribution shapes ('The oldest says,' / '"I thought they died\n
off," says the tallest' / '"No," the old smith said.') and harvests every
designation a speaker is referred to by. Post-filters pronouns, narration
verbs ('Hallr Grips His Axe And'), finite-verb clauses, possessives
('Hallr's' -> 'Hallr', 'Sheyla's voice' -> 'Sheyla'), scene/internal labels,
indefinite collectives, and tags that fall inside a quotation (that is
someone being talked ABOUT, not speaking).
- scripts/merge_designations.py adds any designation the cast does not cover
as an 'unidentified' speaker with its evidence line. Never removes or
rewrites existing entries, atomic tmp+replace, idempotent (0 on re-run).
- scripts/run_discovery.py runs discovery to a SEPARATE output file so the
curated speakers.json is never overwritten by a discovery run.
Result on Last Dwarf: 56 designations, 33 already covered, 23 genuinely
missing — including the ch5 crowd (The Oldest, The Tallest, The Young One,
The One With The Biggest Sword, The Biggest One, The Leader, One Girl Helping)
plus Manfred, The Sergeant, One Witch, The Soldier and others. Cast 70 -> 93,
all new entries flagged 'needs check' at 0.00 confidence.
Also: scripts/run_discovery.py, and a 14-test suite covering the ch5 passage,
possessives, quoted narration, finite-verb clauses and post-quote lowercase
subjects. Full suite 461 passed (8 pre-existing/parallel failures unchanged).
Answers four questions raised about the manifest.
1. The vmatrix IS project data. A voice is defined by description AND vmatrix;
the vmatrix is "the set of numbers inside Qwen3-TTS that IS the voice"
(full ICL ref_code + speaker embedding, x_vector_only=0). Recorded as
AUTHORITATIVE but least-governed:
- keyed by a name-derived slug, not an id, so renames/merges orphan it
- no locale dimension at all, so it cannot be assumed valid for a
translated audiobook; the shared model needs it per (voice, locale)
- coverage is inverted: DealHunter has 38 .pt files, Last Dwarf (active)
has none
- voice_registry.vmatrix_path is empty for all 13 rows, so nothing links
a speaker to its vmatrix
- captured as the canonical record but rendered via embedding mode
(>15 min/utt for ICL vs ~17 s/utt), so constancy is actually enforced
by the marker voice_fp, not the vmatrix
Proposed voice_assets representation added, with source_ref_sha256 as the
link that lets a re-voiced speaker invalidate its vmatrix.
2. The Hallr->HALR respelling IS language-agnostic, verified: apply_phonetics
survives CJK ('Boulanger said' -> 'パン屋 said'). One real gap: ipa_to_simple
ends with a non-ASCII strip, so 'パocht' -> 'OCHT' and 'ɸɯ' -> '' silently.
Explicit spellings work for any script; IPA derivation does not. Documented
as a rule for the aligned definition set.
3. Project lifecycle now recorded, since it changes migration treatment:
DealHunter published (frozen lineage), Last Dwarf active (1,796/520/496
stale), Treasure Island a voice lab (flint_a/b/c variants + a parrot
voice, voice_analysis never run). Requires a lifecycle dimension distinct
from legacy project.status.
4. New section 14 on non-static transition: per-project adoption, additive-only
evolution, definition versioning, vocabulary-as-data, an in-band migration
ledger, idempotent re-import, immutable released lineage, and the audio
path never depending on the shared DB. Section 14.2 lists the definitions
still trapped in code that must be extracted before the schema is frozen.
Also captured analysis/voice_analysis.json as the real measured voice-attribute
dataset: 9 features per voice plus a requirements gate that already includes a
uniqueness check with a nearest_neighbour record.
Inventory of every data object an UmbraAudio project holds, grounded in
read-only introspection of all four live projects on 2026-09-28. Intended as
the input to enterprise database schema design and the transition plan, with
UmbraAudio as the upstream data source for the translation system.
Headline finding: the SQLite databases are mostly empty. Seven of eleven
tables hold zero rows in every project - characters, generated_audio,
pronunciation_guide, quality_reports, review_decisions, voice_review_history
and exception_reviews. Only project, chapters, utterances and name_tokens are
populated. Speaker identity lives solely in speakers.json +
voice_profiles.json + the ref WAVs, and the audio record lives solely on the
filesystem plus the *.wav.txt markers. A migration that reads only SQL would
lose the speaker model, the voice model, the pronunciation model and the
entire audio record.
Also captured: three-way speaker-count disagreement in every project
(Last Dwarf 70/68/71, DealHunter 61/64/94 refs); per-project schema drift
(Treasure Island and Vocals lack name_tokens and nine utterances columns);
four divergent DDL authorities; exact DDL for all tables; JSON key shapes;
artifact naming and the marker-based idempotency contract; data classification
(copyrighted, biometric, personal); a legacy-to-shared-entity mapping table;
and the ten requirements the translation system has of this data, four of
which are blocking.
Verified against the live system on 2026-09-27:
1. The settimeout(None) fix was documented as deployed but was ABSENT from
/Library/LaunchDaemons/qwen_relay.py (mtime Sep 10 06:33), so the bug it
describes - any TTS call over 10s killed mid-response - was still live.
Patched (backup qwen_relay.py.bak.20260927) and restarted the daemon with
launchctl kickstart -k.
2. The relay was never down. launchctl print reports state = running and root
holds 127.0.0.1:18881 LISTEN. An unprivileged lsof cannot see root processes,
which makes it look absent, and curl to a relay with a dead upstream returns
http_code=000 because the relay accepts then drops. 18881 -> 000 means
"relay up, upstream refused", not "relay down". Confirm with sudo lsof.
Actual current failure is the GPU box 192.168.123.14:8881 refusing connections;
its Qwen server is a manual start and is down. qwen_relay.py has no upstream
fallback, so every call is dropped and the app falls through to silent
Edge-TTS. Keeping the relay cannot fix that; local 127.0.0.1:8881 works today
and loopback is exempt from Local Network TCC.
Scene-break rows must carry speaker_name='Narrator'. The system requires a
speaker on every utterance row, so 'scene_break' was never a valid authorial
value - it was a generator artifact. Normalized 74 rows book-wide (81 total).
The value is inert for rendering: is_scene_break() keys off the <<<>>> text and
the scene_break type, never speaker_name, and both generation paths
short-circuit before voice resolution. Verified 62/62 existing scene-break WAVs
are exactly 2.000 s, matching SCENE_BREAK_SILENCE_MS.
Recorded so the transient 'scene_break' speaker value is not mistaken for a
deliberate authorial choice again.
Design-phase contract for moving UmbraAudio + the future UmbraTrans review UI
onto the shared enterprise PostgreSQL database without interrupting audiobook
production. No database contacted, no migration SQL, no production data written.
Verified against the real schema (infra/postgres/001_control_plane.sql, 13
tables, all identity/tenancy/audit) and the full current audio codebase.
Key findings:
- No local accounts exist to migrate; the only identity traces are the literal
string 'user' in two review tables. One canonical enterprise user.
- No durable utterance identity: artifacts are named by mutable
utterance_index (ch{N}/utt_{idx:04d}.wav), which spec 18 forbids as a
durable ID. Proposes a content-addressed source_unit_key recorded at
generation time, leaving on-disk names untouched for continuity.
- BLOCKER: umbracontrol_worker matches zero RLS rows and cannot attribute an
audit event, so worker identities are unrepresentable. Needs
service_identities + job_leases, not a GUC workaround.
- RLS machinery is owned by the postgres superuser because the bootstrap pipes
the schema through psql -U postgres. FORCE ROW LEVEL SECURITY would cause
42P17 infinite recursion; recorded as a load-bearing invariant.
- 10 verified application defects, incl. download traversal via string-prefix
containment and raw upload filenames.
No UmbraAudio behaviour changes. Adapter boundary keeps SQLite+filesystem as
the default repository so cutover is a flag flip and always reversible.
The reviewer found obvious speakers missing from Last Dwarf's cast that had to
be added by hand: the six men in ch5 identified only as 'the oldest', 'the
tallest', 'the young one', 'the one with the biggest sword' and 'the biggest
one', plus the scum group. The prompt had no rule requiring these to be
captured, and nothing stopping narration verbs from being baked into names
(the result: 'Walking tallest', 'Walking oldest', 'Walking Youngest').
- Rule 8 'Every Speaking Line Is a Speaker': a character who speaks any line is
a speaker. A designation ('the tallest', 'the guard', 'a voice', 'the three
scum people') is a valid name. If two designations might be the same person,
still list both and flag the unsure one — a wrong merge is permanent, a
duplicate costs one merge click. Includes the ch5 passage as a worked example
yielding six speakers.
- Rule 9 'Name Hygiene': never include narration verbs/adverbs in a name
('Walking tallest' -> 'the tallest'), never include '(Scene N)' internal
labels, keep capitalisation consistent, no empty names.
- Rule 6 extended: a designation is itself unidentified; never upgrade
'the tallest' into a guessed proper name.
- tests/test_discovery_prompt_rules.py guards all three so a future prompt
rewrite cannot silently drop them.
Re-running discovery is required for the new rules to affect speakers.json;
existing entries are untouched (the designation system flags the current mess
in the meantime). Full suite 447 passed (8 pre-existing/parallel failures).
Addresses two related gaps: discovery misses obvious role/comparative speakers
(the six men in ch5 identified only as 'the oldest', 'the tallest', 'the young
one', 'the one with the biggest sword'), and the speakers page gave no signal
about which speakers are trustworthy.
New umbraaudio/analysis/speaker_confidence.py (deterministic, no LLM) gives every
speaker an identification confidence and a designation of clear / needs_review:
- categorical (blocking) defects that always need a human decision: a narration
verb glued onto the name ('Walking tallest'), an internal '(Scene N)' label, an
uncapitalised name, a name made only of role/comparative words, a probable
unmerged duplicate, discovery 'unidentified', or zero dialogue anywhere
- graded penalties: unknown gender, no traits/description, thin utterance share,
high share of needs_review utterances, unevidenced aliases
- duplicate detection tuned to avoid false merges: 'Lieutenant' vs 'Lieutenant
Stonesplitter' and 'Walking oldest' vs 'Walking Tallest' are NOT duplicates,
while 'Guard 1'/'Guard 2', 'Rielan'/'Rieland' and 'The Girl'/'The Girl helping'
are.
UI: the speakers page now shows all 66 speakers each with a ✓ clear or
⚠ needs check badge; hovering a needs-check badge lists the machine-readable
reasons. A collapsible 'N speakers need checking' panel at the top lists every
flagged speaker with its reasons, plus a header count. Registered as a template
global in create_app so routes.py is untouched (it is being edited concurrently
by another session). Rows render in ~60ms; the assessment is computed once per
page, not per row.
Last Dwarf: 32 clear / 34 need checking, including every speaker the reviewer
had to add by hand (scum leader, scum country boy, Scum guy1, Walking oldest/
Tallest/Youngest/sword, The Girl) and the four real merge pairs.
tests/test_speaker_confidence.py: 17 tests. Full suite 442 passed (8 failures are
the pre-existing/parallel-session voice_profile_manager + test_pronunciation ones).
Two real bugs found by the user clicking Assign on the Name Tokens tab:
1. The action buttons never fired. {{ token|tojson }} emits double quotes
inside onclick="...", so the HTML attribute terminated early
(onclick="ntAssign(8, " and the rest became stray attributes. The browser
threw a JS syntax error, no fetch was ever sent, and the token stayed
unresolved. Switched to data-token/data-idx attributes read in JS, which is
immune to spaces and quotes. Also added real error surfacing (.catch +
server error message) and made the result message explain the 0-utterance
case: a name-tag token like 'Tobias Teppo' comes from narration, not an
assigned speaker, so assigning it records an alias and re-points nothing.
2. The scanner missed a real unnamed speaker. _is_known('Witch') returned True
because single-word containment matched the profile 'The Witch Pair', so the
2 utterances attributed to an unnamed Witch never reached the queue. Expanded
_ROLE_NOUNS with identity/trade/supernatural/machine nouns and made the
single-generic-word guard use it, so generic identity words are never
auto-absorbed into a longer profile. Last Dwarf now reports 10 tokens
(was 9) with 'Witch' (ch40) included.
Verified: real assign round-trips (alias + evidence + utterance re-point),
error paths return 400, tokens with spaces route correctly. Added a regression
test for the Witch containment case. Full suite 425 passed (8 pre-existing /
parallel-session failures unchanged).
Gives the reviewer a single place to resolve every name the manuscript uses as
a speaker that doesn't resolve to a profile, so nothing reaches the translation
export unresolved.
New (separate router module, so the UI never has to merge into the hot
routes.py):
- umbraaudio/web/name_token_routes.py: /name-tokens page, /name-tokens/partial
(HTMX), POST /api/name-tokens/scan, and per-token assign/create/dismiss, plus
/api/speakers/merge-by-name. Registered in create_app.
- templates/name_tokens.html + _name_tokens_list.html: token table with kind,
use count, chapters, evidence, status, and inline Assign / New speaker /
Dismiss actions. Shows a one-click suggestion for name-like tokens.
- base.html nav: Name Tokens + Review (/exception-review) links.
- name_tokens.py: shared scan_project() + build_alias_map() so the CLI and the
web scan use identical logic.
Fixed a pre-existing 500 on /exception-review (now linked in the nav): it called
the old Starlette TemplateResponse(name, context) signature, so the context dict
was used as the template name. Switched to (request, name, context) and replaced
the hardcoded project name DealHunter with the active project.
Verified live: /name-tokens 200 with 8 rows; scan API persisted 9 tokens;
assign error paths return 400; dismiss round-tripped through HTTP with the
scan evidence preserved (Phase 2 fix) and was then reverted. All nav pages 200.
Full suite 424 passed; the 8 failures are the pre-existing/parallel-session
voice_profile_manager + test_pronunciation ones.
routes.py was committed with ONLY the two hunks that are mine (name_tokens
schema + the exception_review fix), via a filtered patch. The parallel session's
5 uncommitted helper_url hunks are left untouched in the working tree.
1. merge_speakers never committed its utterance re-normalization. The CLI
action path also returned without closing the connection, so Python rolled
the transaction back on exit — merges silently lost every utterance rename.
Now commits explicitly, and the CLI closes via try/finally.
Verified: merging Guard 1 -> Guard 2 moved 6 utterances; a fresh connection
sees 0/6 instead of the old 6/0.
2. assign_token_to_speaker wrote a dict into alias_evidence, but the discovery
schema is list[str] and CharacterDiscoverer._alias_has_evidence() calls
.lower() on each item — the next character-discovery run would have raised
AttributeError. Now writes a descriptive sentence that also contains the
alias, so the evidence check returns True for it.
3. dismiss_token overwrote evidence_json with the reason, discarding the scan
evidence. Now preserves the original list and appends 'dismissed: <reason>'.
Added regression tests: cross-connection merge commit (file-backed DB),
alias-evidence string-list schema + discovery compatibility, dismiss evidence
preservation. 46 resolution/scanner tests green; 85 across Phase 1+2+discovery.
Surfaces proper-noun speaker identities in the manuscript that do not resolve to
a known speaker profile, so each can be assigned/created/dismissed/merged by a
reviewer instead of silently mis-attributing.
New:
- umbraaudio/analysis/name_tokens.py: scans two high-signal evidence positions
(unresolved proper-noun speech tags + un-profiled speaker_name values) rather
than every capitalized word (594-token noise -> 9 real orphans on Last Dwarf).
Groups book-wide, offers fuzzy suggestions (word-overlap / difflib / soundex /
4-char common-prefix misspelling), and suppresses personal-name suggestions
for role nouns. Persists to a new name_tokens table (idempotent, preserves
human status across re-runs).
- umbraaudio/analysis/orphan_resolution.py: assign-to-speaker (add alias +
evidence, re-normalize utterances), create-speaker, dismiss, merge-speakers.
Merging is idempotent; matching is case-insensitive.
- scripts/scan_name_tokens.py: scan CLI (--write) + one-shot actions
(--assign/--create/--dismiss/--merge).
- DB: name_tokens table in database.py SCHEMA (web routes create it on demand via
ensure_name_tokens_table, so no shared-file conflict).
Last Dwarf: 9 orphan tokens (8 role speakers + 'Tobias Teppo' name-tag which
suggests 'Tobias'). Merge candidates Guard 1/Guard 2 + Narrator/OrigNarrator are
real speakers flagged for the merge action.
Tests: tests/test_name_tokens.py (30) + tests/test_orphan_resolution.py (12) green.
ruff/mypy clean (TC003 sqlite3/Path notes match the repo's existing pattern in
database.py). Full suite: 420 passed; the 2 test_pronunciation failures and 6
voice_profile_manager accent failures are pre-existing/parallel-session (a
concurrent 'qwen_tts_helper_url' change), unrelated to this work.
NOT committed: umbraaudio/web/routes.py name_tokens schema hunk — that file is
actively being edited by a parallel session (mixed with its helper_url changes);
left in the working tree to avoid a merge hazard. Table is created on demand, so
nothing is functionally blocked.
