Three defects on one card of the step sub-page, all of them a surface
answering a question that belongs to the metric.
"Your usual range is -5,595.6-13,387.6 steps", under a chart whose own
y-axis refuses to draw a negative step. Same app, same question, two
answers: the axis clamped and the sentence beneath it did not. Both now
read the same fact. `VALUE_RANGES` already declares every metric's
plausibility floor for ingest; `value-domain.ts` reads it and answers
"can this metric be negative" for the padded y-domain and for the median
+/- k.MAD band alike. The band's lower edge floors at zero for a
non-negative metric; `spread` stays the unclamped dispersion so a
deviation-in-sigma calculation is unaffected. `buildBaselineBand` now
takes the metric it is describing as a required argument, so a new caller
cannot quietly reintroduce a negative edge. Two types carried no
plausibility range at all, which let any number through ingest and read
as "unknown" on the negativity question: walking steadiness gets the same
0-100 percent band as its two gait siblings, the loud-listening event a
floor of 1 with a generous ceiling.
"13,387.6 steps" also said steps to a reader of German. Five sub-page
units are English words, not symbols - steps, flights, falls, years, and
a "count" that was a poor label in any language. They now travel as i18n
keys through a new `unitKey` on the metric-page scaffold and resolve in
all six locales. The thirteen real symbols (bpm, mmHg, %, m/s, kg/m2,
...) stay literals; translating a symbol would be the opposite mistake.
And the Min read "104.0 steps". Precision was a per-page prop defaulting
to one decimal, so every page had to remember that its metric is a count
- and the step page did not. It follows the metric now: a discrete metric
(tallies, category events, closed integer instrument scores, ordinal
levels) resolves zero decimals wherever it is rendered, and a page only
passes the prop to override that.
`binaryReferenceSex` was already the declared owner of the shrink from the
three stored values to the two a sex-split reference table has rows for, but
eight sites still narrowed inline — including a byte-for-byte private copy in
the effective-range resolver and a ternary 37 lines above a `toProfileSex`
call in the same file. Each of those is a place the fallback can drift.
Delete the copy, replace the ternaries, and name the pair once as
`BinaryReferenceSex` so a signature that needs it stops re-listing the two
literals. No behaviour changes: every site already resolved OTHER and "no
answer" to the same neutral or absent value, and the existing suites pass
untouched.
The properties this provider has to keep: it resolves against the base
URL that was saved for it, the pinned OpenAI arm cannot see that URL and
still posts to api.openai.com, and the OpenAI key never travels to the
gateway. Each is asserted at the wire, not on internals.
The SSRF floor is checked through the same alt-notation classes the
Local provider is checked against, and the pinned tags are shown to keep
requiring a public host even when the operator opened every private one.
The AI settings connect-src is pinned to what it already allowed: the
gateway is called server-side, so no CSP change is needed, and a future
change that needs one will turn this red.
The chain gains a seventh entry. It builds the OpenAI client against the
base URL the user saved, with the bearer they saved for it and the model
they named, and it reads none of the OpenAI columns — the pinned
api.openai.com arm stays pinned.
Host policy is the Local provider's, unchanged: a public host always, a
private one only where the operator allowlisted it, which is what a LAN
LiteLLM or vLLM needs. Egress keeps riding safeFetch.
The JSON-mode dialect the Local client learns per endpoint moves into a
shared module and now serves both clients, so a gateway that rejects
response_format is detected once and retried without it.
The entry stays out of the default chain: it can only resolve against a
base URL somebody configured.
The preamble already says once what the two states mean and that neither is
absence; repeating it per row would have doubled a manifest that rides every
turn.
The unit suite mocks Prisma, so it proves the discriminant and nothing about the
query underneath it — a wrong column would have left every test green and every
account answering no_data_unconfirmed in production. Seeds the reported case and
the four other shapes: an empty record, another account's rows, tombstoned rows,
and a history that mixes units.
Rounds the out-of-window mean to one decimal, like every other mean the snapshot
surfaces, and annotates a failed manifest probe rather than letting it degrade in
silence.
The mutation check found both holes: an addendum test that only looked for the
new reason codes passed with the old declare-absence-and-move-on sentence still
in place, and a block-failure check whose window was wide enough to see the
neighbouring block's annotation.
The fixture is the report: a fortnight of glucose readings from April 2024 on an
account that displays them everywhere else. The clock is pinned so it cannot age
out of the window it was written to sit outside of.
Covers the four states an empty window read can resolve to, that a never-recorded
metric and a historical one differ in the SAME run, that a mean is withheld
across mixed units rather than fabricated, that the manifest advertises the range
instead of "absent", that both instructions branch on the reason code, and that a
miss carrying figures reaches the prose verifier's authoritative set. Adds the
medication-compliance page to the failed-read sweep and pins the narration
blocks' failure annotation.
A Coach retrieval tool that read the window and found nothing returned
`reason: "no_data"` — the same code for two situations that have nothing in
common: the record holds nothing for that metric, and the record holds a
substantial history for it, all of it older than the window that was searched.
Nothing downstream could tell them apart, so the Coach said the honest-sounding
sentence for the wrong case, and then, having been told there was nothing there,
changed the subject. Someone with a year of imported glucose readings was told
the readings did not exist, on the same account whose glucose page was showing
them.
The window cap stays. It exists because a multi-year history does not fit a
prompt budget, and widening it would only move the same failure. What was
missing is the ability to say a different true sentence.
- `tools/availability.ts` — one grouped aggregate per miss, no rows
materialised, resolving four states: `no_data` (probe confirmed empty),
`outside_window` (rows exist, older than the searched window; carries the
count, the date range in the user's own timezone, and a bounded per-series
mean/min/max), `unavailable_in_scope` (rows exist inside the window and the
domain still produced none — a module is off), and `no_data_unconfirmed` (the
probe itself failed; absence was never established, so it is not asserted).
- Every empty read in the executor routes through it. A bare `no_data` can no
longer be returned without the record having been checked.
- The DATA INVENTORY carries the same distinction, so a domain the window
cannot reach is advertised with what it holds instead of as absent — the model
no longer skips the fetch, and a vague follow-up keeps the subject.
- Both instructions that told the model to report absence and move on now
branch on the reason code, and forbid answering about one metric by reporting
another.
- The out-of-window aggregate joins the authoritative set for the prose
verifier, so the sentence the fix exists to make sayable is not elided.
- A cross-cutting narration block that throws is now counted rather than
vanishing into a `null` that looks like honest absence.
- The medication-compliance page tells a failed read apart from an account with
no medications, instead of inviting someone to re-enter what it already holds.
Refs #648
Adds the cases that were missing everywhere except the FHIR conformance
suite: an empty string clears without taking its siblings down, OTHER
round-trips through the profile and onboarding writes, an unsupported
value is still refused and says which field it is about, the Strain
model scores an unanswered account on neither side, the dashboard bands
give OTHER the neutral fallback, the snapshot payload carries the value
instead of flattening it, and the doctor-report cover omits the row
rather than claiming a value the account never chose.
Refs #645
Mood has been a first-class module here for a long time, with entries,
structured tags, rated factors and notes. The bridge to the external moodLog
service had nothing left to add, and its Settings card was withdrawn in
a56e6fe61 on the promise of a dedicated removal. This delivers it.
Removed: the hourly `moodlog-sync` queue, cron and handler binding; the
inbound webhook and its shared-secret handling; the reverse push and its four
call sites; the pull-sync module; the per-user credential routes and the
instance-wide toggle; the entry on the `/api/integrations/status` envelope,
the `IntegrationKey` union, the cadence map and the display-name table; the
boot-time secret rotation pass; 22 message keys per locale.
Migration 0271 drops the five `users` credential columns and
`app_settings.mood_log_global`, clears the ledger row for the provider, and
deletes the retired pg-boss schedule so no cron keeps enqueuing work no
handler drains.
Mood data is untouched. `mood_entries` is not named anywhere in the migration
and cannot be reached from it: the dropped columns carry no index, constraint
or foreign key, and `mood_entries.user_id` references `users(id)`, not any of
them. Rows the bridge imported keep `source = 'MOODLOG'`, and every read path
- list filter, source badge, exports, backups - still resolves that value.
`src/lib/validations/moodlog.ts` held the native mood-entry schemas alongside
the bridge's own; it is now `validations/mood.ts` and carries only the former.
Route the medication-compliance status card through finalizeStatusSummary so
it enforces the same dose / risk / causal outbound contracts its seven sibling
status cards do, not only the unparseable-envelope withhold it already had. A
dose-change imperative or a fabricated figure is now withheld to the controlled
fallback on this surface too.
Add the status cards' structured-envelope detection to the two background hero
lines (arrival reaction, coach nudge): a brace-led or fenced JSON reply, or a
truncated envelope under the char cap, now resolves to the deterministic lead
rather than rendering raw.
Fold in the residuals from the same sweep: strip a truncated SUGGEST-ACTION
marker from the Coach prose tail, drop an unreachable zero-rate branch in the
adherence fold, and document why the internal conversation summary stays off
the render-boundary screen.
Replace the internal "[unverified]" token the grounding guard left in a Coach
reply with the editorial elision mark […], and add a quiet muted caption
under the bubble explaining that some figures could not be checked against the
user's data. The mark holds the slot grammar-safely mid-sentence, before a
surviving unit, and inside a range; the notice carries the meaning. The withheld
count rides the existing provenance envelope (unverifiedFigures) through the
stream, persistence, and reload, and the fenced document-chat path carries a
minimal envelope so its notice survives too. New i18n key across all six locales.
Also fix an adjacent reload gap: provenanceFromJson restored only a subset of the
serialised envelope, so a persisted suggestion card, confirm action card, and the
tool trace were dropped on conversation reload despite the render layer expecting
them to survive. Restore them defensively alongside the new field.
Promote the per-turn magnitude bag to a per-conversation typed Grounding
Ledger built from every source the model was actually shown: this turn's
tool payloads, prior-turn tool figures, coach-memory goals, the medication
schedule, the reference-grounding block, and guided blocks. Assistant prose
is never a registration source, so a figure that slipped a rung cannot
launder into an authoritative value on the next turn.
Complete the action ladder: a reconciled number passes, a reference or
education sentence is exempt across all six locales, a positively-typed
benign number (date, time, year, ordinal, list marker) passes while an
untyped unreconciled one stays on the strip path, and an unreconciled
health figure is soft-stripped.
Gate the dose continuation exemption on the user's actual schedule: a
"keep taking your N mg" is trusted only when N is a dose the user is on;
an off-schedule maintenance dose is caught. Extend the risk and percent
classifiers to fr/es/it/pl to match the EN/DE depth.
No schema migration: the ledger is rebuilt per turn from persisted tool
traces and re-rendered prompt blocks.
Replace the untyped magnitude bag in the Coach numeric verifier with a typed
tokenizer and a normal-form reconciler, and narrow the outbound risk and dose
banks so correct figures and correct refusals stop being blocked while every
fabrication shape still blocks.
Grounding verifier (coach-prose-grounding.ts):
- Classify each prose number before grading — ISO/written dates, clock times,
bare years, ordinals, list markers, hyphenated time spans, ranges, percents,
thousands-separated integers, unit-suffixed measurements. Benign types are
never graded or stripped; ranges decompose into per-endpoint magnitudes.
- Grade against typed authoritative entries (value + kind + aggregation).
Sign-insensitive, canonical rounding, ratio to percent in percent space, and
minutes to hours widen the match only for an aggregate, kinded entry; kind
scoping stops a weight value grounding a blood-pressure claim.
- Thousands/decimal separators follow the reply language.
- Boundary-safe strip so a flagged token never clips a larger grounded number.
Outbound screen (outbound-screen.ts):
- Risk bank requires an asserted figure or a categorical engine result; a bare
engine or horizon mention (education/refusal) passes. Spelled-out percents
count as figures. No refusal-context exemption. EN + DE.
- Dose bank keeps a maintenance-continuation restatement, voided by any change
stem. Taper suffix fixed.
Route + fenced-chat thread the reply locale into the verifier.
Merge the eight reviewed audit fixes that were cut a week ago and never
landed. Main has moved on since, so each fix was re-checked against the
current tree and each of the five conflicts resolved by understanding
both sides.
Kept, still open on main:
- Proxy early exits now carry the transport security header subset via a
shared helper. Reconciled with main's /api/health worker-only carve-out.
- Batch measurement ingest bounds the entry instant so a future-dated row
can no longer win "latest reading" or poison the rollup buckets.
- Web Push dials through the pinned safeFetch dispatcher with
requirePublicHost, closing the last user-controlled-host egress.
- Mood bulk upsert and its tag/factor links run in one transaction, so a
rejected rating rolls the mood row back instead of leaving a tagless row.
- Full backup carries NutrientIntakeDay with its unit and source. Folded
into main's portable-export / disaster-recovery split.
- Live daily-series fallback restores the min/max spread band the tiered
read already emitted.
- Coach stream aborts its in-flight reader on unmount and resets send state
on a conversation switch.
- Manual water writes share the day-bound helper, nutrient add clamps after
the increment, and the thiamin cap drops to a real supplement ceiling.
- Compliance point window widens symmetrically instead of wrapping overnight.
- Oura, Polar and Fitbit token helpers classify an empty or non-JSON 2xx
body as a handled transient error.
- WebAuthn relying-party fallback to localhost is gated to development.
- Chart-token strip list is derived from the allowlist and pinned by a test.
- Regression tests pin the snapshot-cache tenant key, the GlitchTip
path-secret redaction and the Withings credentials delete.
Superseded by main, tests re-pointed at the current shape:
- The intake CSV import was rewritten on main to run through a background
job with a single-active-job admission gate. The future-date bound is
re-applied at request validation and the per-user limiter is kept for
parity with the bulk twin; the old synchronous-write test is replaced.
The previous commit spliced the DATA INVENTORY manifest into the tool-mode
authoritative payload set unconditionally. That made the payload set
non-empty on a turn where the model answered WITHOUT calling any tool
(empty tool results), which flipped the verifier from dormant to active —
grading the reply against an inventory that carries per-domain sample counts
only, never the snapshot's pre-computed averages. A figure the model
correctly narrated from the snapshot (a 30-day systolic average) was then
stripped to "[unverified]" — the exact false positive this work exists to
prevent, and a regression from main's "no tool figures this turn → skip,
prompt-level grounding is the backstop" contract.
The inventory must only WIDEN an already-active authoritative set, never
ACTIVATE it on its own. Gate activation on a real present figure block
(a pinned workout-evidence block or a present tool result) exactly as main
did; when active, the inventory counts still join the set so a plain count
restatement stays grounded.
Adds a route-level regression test (fast suite) proving the verifier stays
dormant when no tool returned figures, plus companion unit tests documenting
that a real tool section grounds the average while a counts-only inventory
does not.
Three confirmed false-positive shapes in findUnverifiedCoachNumbers/
stripUnverifiedNumbers were replacing legitimate prose with the literal
"[unverified]" placeholder even though the underlying data was grounded:
- A correctly-computed delta, midpoint average, or percent change between
two of a payload's own headline aggregate figures (e.g. "down 3.5 kg from
your 7-day average") was never itself a raw leaf in any tool result, so it
was flagged as invented. The authoritative set now also carries the
pairwise delta/average/percent-change between a SINGLE payload's own
named aggregate fields (latest, avg7, avg30, avgSys30, allTimeAvg, ...),
scoped per payload so a weight delta cannot ground itself against an
unrelated metric's figure.
- The always-sent DATA INVENTORY manifest (tool-mode base context) carries a
sample count per domain, but that count was never added to the
authoritative set the way the pinned workout-evidence block already is —
so a plain restatement of a domain's count ("you've logged 42 BP
readings") was flagged whenever the model didn't also call that domain's
retrieval tool in the same turn. The inventory entries now splice into
toolResultPayloads exactly like workoutEvidence.
- A calendar-date ordinal ("July 21st") or a numbered-list marker ("1. Cut
back on sodium") tokenised as a bare number and, when ungrounded, got
soft-stripped mid-word into visibly broken text ("July [unverified]st").
Prose is now masked for these two narrow, high-precision shapes before
numeric extraction; the original text used for stripping is untouched.
A genuinely fabricated figure (unrelated to any grounded point, inventory
count, date, or list marker) is still caught and stripped — covered by new
adversarial cases alongside each fix.
Investigated and ruled out: multi-round tool-loop result accumulation (all
present results across every round already reach the verifier); unit
conversion (the app resolves and serves one consistent unit, so there is no
independent conversion step for the model to get wrong). The Daily
Briefing's separate withhold-a-whole-section grounding gate is unrelated to
the reported symptom (a different failure mode, whole-line omission vs.
inline replacement) and was not touched.
The named-risk-engine pattern in the outbound safety screen used
score2? (the 2 optional), so it matched bare "score" too — any mention
of one of this app's own computed, documented scores (Sleep Score,
Readiness Score, Health Score, Recovery Score, Stress Score, Strain
Score) tripped the bank meant only to catch a fabricated named
clinical-risk calculator (SCORE2, Framingham, ASCVD, QRISK).
Require the literal "SCORE2" instead. The other risk patterns —
percentage-shaped risk claims and the other three named engines — are
untouched, so a genuinely fabricated risk figure or a named
risk-calculator claim still blocks.
Closes#587.
The harness landed in v1.30.35, after the slices that built this
release were already cut — so the reaction line's prompt module was
never registered, and its own docblock's warning against the
two-locale collapse was untested. Registering it surfaced a real gap:
the meaning-first instruction lived only in the system prompt's own
words, never in the user prompt in the family's shared phrasing, which
is where the rest of the surfaces carry it and where the tone contract
checks for it. The German body carried no such instruction at all.
Threads locale through the user-prompt builder so the fixed
instruction line matches the reader's own reply language, and wires
the per-user per-day opener-archetype rotation the milestone spec
called for and the first implementation had not connected.
Four suites, each mutation-checked against a reverted protection:
- route-workout-scope: the workout row is read exactly once across a ten-turn
conversation that keeps sending the id, the where clause always names
userId, another user's id resolves to nothing and the chat proceeds, and
the section sits inside the health-data fence.
- workout-evidence: the numbers-only closure. An instruction-shaped
sport_type — reachable through the restore path, which types the column as
free text — folds to "other" and never reaches the block; a free-text leaf
at any depth throws.
- page-launch-params: an unknown ?scope= is ignored silently, and ?workout=
yields to both ?c= and the fenced ?doc= path.
- composite-score-anatomy-coach: every score sheet carries an opener, a
scope narrowed to the composite's inputs, and auto-send.
resolveLaunchState's shape assertion gains the workoutId field.
The spine dispatches a generator-free enqueue; the worker resolves the
provider. Five gates stand in front of it in ascending cost order — both
modules on, ten-minute duration floor, a hard count of four a day in the
user's own timezone, an input hash over the deterministic evidence, and the
token ledger inside the shared status chokepoint. Each refusal returns a
status rather than throwing, because retrying against a ceiling that does not
move until midnight is an unbounded loop.
The detail route only reads: a workout with no row serves null and renders no
card, which is every workout that predates this and every re-synced one. An
undecryptable paragraph degrades to no card rather than failing the page.
Today gains two additive-nullable digest fields. `justIn` names the day's
newest arrival while it is still news — under three hours old, after which the
record is simply current rather than new. `reactionLine` carries a single
generated sentence about that arrival and replaces the hero's standing lead for
the rest of the local day, never appending a second paragraph, so the hero stays
one lead line tall and nothing below it shifts.
Nothing here polls. The open dashboard picks the change up on the cadence it
already runs — a state change, not a notification.
The generated sentence is bounded before it is written: the arrival spine's
unique row per kind per local day IS the throttle, the worker re-checks the
same claim before spending, and the reservation is reconciled on every exit
including each failure path. Without a provider, without consent, or against an
exhausted budget the row stays line-less and the surface falls back to the
deterministic lead — the chip and the provisional-to-final flip are unaffected.
The chip's time string is formatted client-side after mount; `justIn.at`
crosses the wire as an ISO instant so the server never renders a local
wall-clock time the browser would disagree with.
Status cards and derived-score sheets read one way: they state a verdict and
end there. The metric pages already carried the hand-off; the score sheets
passed neither an opener nor a scope, so they painted no action at all.
- Score sheets (sleep / readiness / recovery / stress / strain) gain the same
auto-sent opener the metric pages use, each scoped to the inputs that drive
the composite rather than to a block it has no equivalent of.
- Workout detail's coach launch carries the workout itself. The chat request
gains an optional workoutId; on the FIRST turn the route narrows by
{ id, userId } and pins ONE bounded evidence section onto the snapshot.
Later turns reuse the pinned snapshot, so per-turn work stays constant in
conversation length.
- /coach?workout=<id> covers the full-page and maximize paths.
The workout block deliberately does not take the fenced transport the stored-
document path takes: a workout row is the user's own sensor record, and the
projection admits no free text. That is enforced, not assumed — sport_type is
a TEXT column, so it is re-asserted against the closed sport union, metadata
is excluded apart from its numeric zone durations read through a Zod slice,
and a recursive guard throws on any non-numeric leaf.
Own-history sport averages move to src/lib/workouts/sport-context.ts so the
narrative and the detail page cannot quote different figures.
The evidence block is a numbers-only server projection: zone seconds, the
workout-window heart-rate shape, route climb and a same-sport 90-day own-history
median. Free-text metadata never enters it, and the sport string is narrowed
against the closed enum first, so the payload carries no text the user did not
generate through a measured field.
The prompt composes the shared assessment base and adds the session contract:
device-attributed, own-history-comparative, effort acknowledged where the
numbers show it, and no training prescriptions of any kind. Registered in the
tone harness beside the assessment cards.
Each of these was already correct in the tree and had nothing asserting it, so
a refactor would have regressed it silently.
The decrypted-snapshot LRU keys on userId, but only same-user memoisation and
scope-change recomputation were covered — dropping the userId segment would have
served one account's health snapshot to another with a green suite. The project
has shipped one cross-user cache-leak fix already; this is a repeat class.
The GlitchTip forwarder redacts path-segment secrets in the url it sends. The
query-string strip was tested; the path-segment pass was not, and that is the
one that reaches a durable external incident UI.
The Withings credentials DELETE narrows its connection-row delete to P2025 and
rethrows everything else, so a real failure can no longer null the client
credentials and report a successful disconnect while the encrypted OAuth tokens
survive. The route had no test file at all.
The no-signal floors were the coldest text the product renders. They run when
there is nothing to ground against, and they opened on a bare order — "Measure
blood pressure at rest and under comparable conditions", "Evaluate mood trends
over several weeks" — so the one moment the surface has nothing to say was also
the one moment it told the reader what to do.
They now open on what the surface can honestly say: no assessment is available
right now. That is the meaning-first opening for this case, and it is the only
one available, because a floor with no signal may not imply a read. What
follows describes how the measure behaves rather than ordering the reader
around. The signal-grounded lines above them are unchanged.
The prompt version bump re-fingerprints the assessments cached under the old
biomarker prompt; the per-card input hash folds the constant in, so the
regeneration is one-time and paced by the nightly warm.
The German prompt-injection test asserted only inside `if (decision.refuse)`,
so it stayed green whether or not the filter fired. Un-guard it, drive it from a
case table, and widen the two German patterns to cover the determiners that
walked straight past them ("saemtliche", "die", "deine").
The Oura / Polar / Fitbit token helpers cast a possibly-null `json` to the token
type on a status-only success gate. A 2xx with an empty or non-JSON body handed
callers a null they dereferenced, surfacing an unclassified TypeError instead of
a handled integration error. Narrow explicitly and classify as transient.
The per-biomarker assessment was the last surface in the family building its
prompt inline: a five-line scaffold that skipped the base system prompt, took
no opener hint, and instructed the model to "state the current value" — so the
one card where a bare number reads most like a verdict was value-first by
instruction while every sibling had moved to meaning-first.
It now composes exactly as the seven bespoke cards do. The new prompt module
carries a METRIC section that states the one thing this card needs and the
others do not: its snapshot has no signal block, so the read comes from the
pre-computed range placement plus the direction across draws.
Also point both archetype-driven cards at a registered budget. The generic
metric card and the biomarker card each carried temperature/maxTokens inline
at 0.45/1000, so the registry that documents every surface's ceiling was
silently wrong about the two widest assessment surfaces.
The single-marker no-key fallback replaces the multi-metric overview tip the
card used to render — text about watching several metrics at once, on a card
showing one lab value.
Three surfaces where the words no longer matched the code, plus the missing
web control for a preference the server already honours.
Delete confirmations. Four dialogs promised permanence on surfaces that
render an Undo: the measurements and mood bulk bodies, the lab-result body,
and the illness-episode body (which said "cannot be undone" sixty lines
below the code that fires the Undo toast). All four now use the same
sentence the single-row measurement and mood dialogs already used. A sweep
of every other confirmation in the tree found the rest accurate — the
biomarker cascade, the mood-tag purge, and the account purge keep their
warning because those deletes really are unrecoverable.
Sleep debt. v1.19.0 replaced the summed shortfall with a running balance
that credits catch-up sleep; the labels still described the old sum, so the
word contradicted the number for anyone who slept in after a short night.
Corrected in all six locales, plus the places the old semantics had
propagated: the Coach system prompt, the OpenAPI field descriptions the
native client reads, and four module docblocks.
Glucose reference band. `hasDiabetes` selects the tighter reference band and
grounds the Coach, but no web surface set it — a web-only account was pinned
to the non-diabetic scale with no way out. The control goes on Targets,
above the per-metric editor, because it answers the same question that page
answers and the two are ordered: it picks the default band, an override
below beats it. The copy says what it is — a reference band, not a
diagnosis, never inferred from a reading, no clinical action taken on it.
The passkey list was reported as always reading "never used". It does not
reproduce: the column is selected by the list route and stamped on every
verified assertion, both since v1.23.0. Pinned with tests on both halves so
the contract cannot quietly regress.
The retry suffix still described the strict insight wrapper that was removed
long ago: it demanded citations[] and warnings[] — fields insightResultSchema
does not have — required metricSource, rationale, id and severity on every
recommendation when all four are optional, and named none of the five fields
the live schema actually requires (classification, findings, correlations,
dataQuality, disclaimer).
So the one corrective pass steered the model away from the contract it was
being corrected towards. A reply that obeyed the correction could not
validate, fell through parseComprehensiveResult's raw-object escape hatch,
and the downstream surfaces read an unvalidated payload.
Describe insightResultSchema and nothing else, and carry the contract as data
in RETRY_CONTRACT_EXAMPLE so a test can assert prompt and validator agree.
The Today hero stayed on its all-clear line while a dose was actually due:
the dose-window item only fired once the slot was already overdue, so a
dose the user could take right now produced no rail item at all. Emit the
item from the moment the dose becomes takeable — the on-time window's lead
plus the early grace, the same instant logging it starts crediting the
slot — and give the two faces distinct wording and status: a due dose reads
calm and informational, an overdue one keeps the warning. A settled day
still emits nothing, and a cached block whose anchor has since passed keeps
rendering neither face. Push reminders are untouched; they still fire only
on overdue.
The briefing grounding gate discarded the whole block over one number it
could not verify, and it could not verify plenty of legitimate ones: the
allow-set was a hand-maintained list of feature blocks that had already
drifted from the prompt three times, and still omitted medication
compliance, the context counters, the correlation coefficients, every
coverage object and the comparison snapshot — all of which the prompt hands
the model verbatim. Derive the allow-set by walking the prompt payload
instead, so grounded means exactly what it says: a figure the server gave
the model. A fabricated number still matches nothing.
Fix the disposal to match. A content-failed retry left a hole while a
transport-failed one fell back to the previous cached briefing, though in
both cases the rejected text is discarded and the previous briefing passed
this same gate when it was written. Fall back on both paths, and only
report the briefing as withheld when there is genuinely nothing to show.
The on-demand route had no fallback at all; it now mirrors the background
generator that writes the same cached payload.
Four locales received English on surfaces whose translations were
already complete, because the resolving code was a de/en binary.
- Medication reminder push and the Telegram inline-keyboard labels
admitted only de/en, so es/fr/it/pl got the English template on every
channel. Route the resolver through a shared coerceLocale().
- The Coach's trailing "reply now in X" clause said English for any
non-German locale, overriding the correct system-prompt directive.
Read the language name from a shared map instead.
- Coach refusal copy was a de/en constant pair, answering four locales
in English at the one turn the Coach is declining to help. Move the
copy into the bundles; the EN/DE constants now derive from them.
- The sleep-stage chart formatted axis ticks as en-US for four locales;
the cycle calendar and the device-score tile followed the *browser*
locale rather than the app locale.
- Three components carried a byte-identical duration formatter that
branched on German. Lift it to one key-driven helper.
Polish needs a third plural tier: 2-4 takes a "few" form a two-form
scheme cannot express, so "2 godzin temu" read as "2 of-hours ago".
Add a pluralKey() helper driven by Intl.PluralRules and a Few tier to
the relative-time and stale-hint keys. The five locales whose rules
have no "few" category resolve exactly as before by construction, and
their Few values mirror Other. Teach the reverse-coverage guard to
expand a pluralKey() base into its three tiers.
generateInsight() parsed against the strict aiInsightResponseSchema,
cross-checked that every recommendation's metricSource appeared in
citations[], and threw InsightSchemaError on a second failure. Its only
caller was runWithFallback(), which had no production caller — every
live surface runs runRawCompletionWithFallback. So none of it executed.
Wiring it into the live path was the alternative and it is not viable:
the live insightResultSchema has no citations[] field at all and treats
metricSource as optional, so the cross-check would have rejected
essentially every generation rather than tightening anything.
Removed: runWithFallback and its param/result types, generateInsight
and its private helpers, findUncitedRecommendations,
findRecommendationsMissingRationale, the never-thrown InsightSchemaError
and the isHardProviderFailure branch that excluded it, and
computeConfidence — a deterministic override that only ran inside the
dead wrapper.
Kept: buildRetryCorrectionMessage, which comprehensive-generate sends on
its own retry leg; the file it lived in is renamed to match what it now
holds. aiInsightResponseSchema stays as the shape that message
describes. Note the retry prompt asks for a stricter payload than the
live validator requires — that mismatch predates this and is left alone.
computeCitationCoverage is untouched and still live: it is a metric, not
a gate, and v1.30.24 wired it into the comprehensive post-parse step.
Chain-runner coverage the deleted tests carried (last-working cache,
empty-input guard) is ported onto runRawCompletionWithFallback, which
shares the machinery.
Three things the write surface claimed but did not hold.
Scope: the run bodies never re-checked ctx.canWrite. Enforcement was
real but lived entirely in the registration decision, and the test
asserted registration only — so a refactor that registered the write
tools unconditionally and meant to refuse at call time would open the
surface with the suite still green. Wrap every run body in a scope
guard at the registry level, so a new write tool inherits it, and drive
the run bodies directly from a read-only context in the test.
Free text: medication names, doses, treatment classes and lab analyte
names were interpolated into tool prose with no delimiter. A lab
analyte name is transcribed out of an uploaded document, so a hostile
document chooses it, and the write tools ride the same wire. Fence the
assembled prose with the shared data-fence markers and scrub short
result titles, which would leak markers into a host's result list.
Confirm: the header called the confirm flag a PROTOCOL-LEVEL CONFIRM
GATE. It is a client-asserted boolean with nothing binding it to a
prior preview. Binding it to a server-issued preview token would prove
only that a preview call happened, which any client can do, so it buys
no human-in-the-loop for a round trip and server state the stateless
transport does not want. Describe it as the convention it is and name
what actually bounds the blast radius.
Marc-André Bombeck