Google Street View registers beside Mapillary as provider `google`. It runs
on the Maps JavaScript API with the browser GOOGLE_MAPS_API_KEY, loaded on
first use (mapsLoader.js), and uses Google's own StreetViewPanorama as its
viewer adapter, emitting provider-neutral poses (address and month in the
caption, a link to Google Maps, steps along Google's arrows, FOLLOW on
Google 3D). Its chip starts off because every panorama is billed to the
key; the share option is `g`.
Google offers no coverage layer for other maps, so the provider draws
none: with it on, a click on bare ground at street zoom opens its nearest
Google-collected outdoor panorama (groundClick.js). A key Google refuses
reads KEY REJECTED and fails a pending open at once; a lookup Google never
answers fails after 10 s.
Core: providers may declare `groundClick` and `defaultOn`; poses may carry
`capturedAtPrecision` and `title`; the coverage line keeps every
provider's hint; opening one provider's image releases another's
prewarmed viewer; FIT/FILL hides for viewers without render modes. CSP
script-src admits https://maps.googleapis.com. The browser gate gains two
Street View steps (the keyed one skips as 'no Google key' in CI).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The browser gate is one pass over the remaining flows: rail strip and
expand, coverage with fixtures, the ON/OFF pill, 360° and SINCE filters,
KEY REJECTED, nearest photo with caption, × close, phone fit, keyless
KEY REQUIRED on a second page and no page errors. Panel controls are
pressed through the DOM; pick-point search, line/cone clicks, the
framing-cancel, reload/share and status-route probes, the second
viewport and pointer-settling retries are gone (unit tests cover them).
Tests that pinned constants, copy, CSS or source text, or repeated
another test, are removed or merged. CSS loses an unused variable and
duplicate rules. Docs describe the slimmed feature; unused exports and
the layer's dead setCoverageFilter/refreshCoverage surface are removed.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Drop the disk tile cache (.gev-cache/mapillary/tiles, background sweep,
size cap, legacy retrim), the server-side 401/403/429 holds, the upstream
concurrency semaphore and per-waiter abort bookkeeping. The client already
pauses on 429 and stops on a rejected key.
Keep: server-side token, {configured}-only status, strict coverage path
validation (z11-14), image-layer trimming, same-site gate and per-IP limit,
HTTPS redirect pinning to the tile origin, the streamed size cap, the error
mapping the client reads, a 24 h byte-budget LRU memory cache, in-flight
coalescing via the shared coalesceProxyRequest, and browser Cache-Control.
Routes and constants fold into mapillary.js and mapillary/tiles.js.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The layer now owns its Mapillary parts directly: the provider registry,
pick router, multi-provider UI summary, params codec helper and provider
context are gone. createStreetLevelLayer({ source }) builds coverage,
sequences, the nearest-image lookup and the MapillaryJS viewer itself; a
single providerOn switch keeps the share link's `mapillary` option (off
draws nothing), and getParams/setParams keep { mapillary, pano, sinceDays }.
No z0-5 overview points: above 60 km nothing is drawn and the panel says to
zoom in; the proxy now serves only coverage z11-14. Tiles come from the
visible ground within range of the camera, ranked from the ground under it
(no line-of-sight ranking or near-range padding). MapillaryJS loads on the
first photo; viewer prewarm is removed. Key status, a rejected key and a
rate limit share one `blocked` state on coverage.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
FOLLOW is gone: no setFollow, follow state, map-stack availability or
handoff listener. Opening a photo still frames it with one flight through
the application's deferred navigation (begin/reassert), so a slow photo
never overrides newer navigation and tracking is released as before.
Ground casting is gone: groundCast.js, meshSampler.js, the surface mode,
the draped-to-cast swap and remesh. Coverage lines are always draped
GroundPolylinePrimitives (ClassificationType.BOTH); cones and the marker
clamp to the ground and keep the horizon cull. The terrain/meshFloor/ground
services and meshFloor.visibleTilesetLoaded are no longer needed.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
EXPAND now uses the browser's Fullscreen API on the viewer element, which
stays in the panel; the hand-built modal (re-parenting to <body>, inert
siblings, focus trap, document-level Esc/Tab, Clean View/recording/cockpit
hide rules) is gone. The button is hidden where the API is missing.
The FOLLOW button and the PROVIDERS chip row are removed. The header pill is
the only layer switch and re-enables a provider switched off elsewhere; under
KEY REQUIRED the error line says which key to add. Chip colour support in
chipGroup, dead CSS and tests pinning the removed behaviour go with them.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
PR #768 bundled a portable-panel windowing system with the Street Level
layer. Remove it so CCTV and Street Level are ordinary docked right-rail
panels again, as on main before the merge.
Removed:
- header drag to float, 8-way edge/corner resize, double-press snap-back,
dockOnCollapse, command-dock clearance, float hint toast, floating z floor,
persisted floating geometry and ResizeObserver re-clamping
(panelPositionControls.js restored to its pre-#768 version)
- src/ui/panelResize.js and its tests, panelDock.test.mjs
- onPanelResized/dockPanel wiring in PanelChrome, the application shell and
StreetLevelControls (EXPAND/SHRINK no longer docks)
- :not(.panel-floating) guards and .panel-floating/.panel-portable/resize CSS
- qa:panel-resize, qa-panel-resize.mjs, qa-panelDrag.mjs, their tests, the
CI "Portable panel gate" step, QA_FAIL_ON_RETRY, and the Street Level
gate's float/resize/dock steps
- floating-panel docs in README, CHANGELOG, CURRENT-STATE, KNOWN-ISSUES
Kept: the Street Level panel as a docked rail panel (share/cockpit/rail
lists, layout controller, rail and Cyber CSS, EXPAND viewer), and the
independent right-rail alignment fix (settledLeftRailTop, 9dde7b4).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Trim the doc text this PR adds (README, CHANGELOG, CURRENT-STATE,
KNOWN-ISSUES, SECURITY, DATA_SOURCES, .env.example, the CI job comment),
drop code comments that restate the code, cut design-note JSDoc to a line
or two, and drop review tags from test names. Comments only; no code
changes.
Also: README key count (seven keys, two metered), CURRENT-STATE's gate
paragraph (CI runs a production build with a dummy token), the stacked
doc above meshBoxInRange and the orphaned inMeshRange line in
coverage.js, and a duplicate ALPR row in DATA_SOURCES that a merge had
brought back.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
CI on fb024d5 failed step 24 with QA_FAIL_ON_RETRY=1: the press after a
double-click dock landed on the globe because the panel, showing a photo
that was still settling, moved 116 px up and lost 46 px of height between
the pre-press hit check and the press (box 1058,382 330x471 -> 1058,266
330x425, from the new diagnostics and the uploaded screenshot). That is
the panel's own content changing its layout, not the rail race fixed in
9dde7b4.
- Fail-on-retry now fails a miss only when the panel's box was the same
at the hit check and at the press (the pointer missed a header that
stood still); a miss the panel moving explains is retried with a
::warning:: naming the move. New harness test with a moving panel.
- Step 24 waits for the open photo and its sequence to finish loading
before pressing the header again.
Verified: strict fixture gate on a production build 28 ok + 2 "no Google
3D" skips and the panel gate 6/6, both with QA_FAIL_ON_RETRY=1.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- The CI gate builds the app and serves it with `vite preview`, with a
dummy non-secret client token (a fork PR never gets secrets), and
answers the Graph API, image tiles and thumbnails from fixtures: 41
photos on a line under the parked camera and one generated 640x320 JPEG
(scripts/fixtures/street-level). All 7 photo steps now run in CI, plus a
real click on a coverage line and on a cone, the framing flight stopping
on close, the inert modal, the viewer returning into the panel, and a
reload keeping a collapsed panel collapsed (share ui token t).
- Fixture interception is limited to Mapillary hosts via CDP
Fetch.enable: page-wide request interception stalled Cesium's workers,
so the old fixture mode never drew anything on the globe.
- Failure evidence (screenshots, UI state, console, failed requests,
render errors) is written to qa-artifacts/ with keys and tokens redacted
and uploaded on failure; qa-artifacts/ is gitignored.
- Cesium render-loop errors fail the gate; Google 3D branches that cannot
run print as skips, and --strict rejects any other skip.
- Panel presses log their diagnostics, warn on a retry, and CI now fails
on any retry (QA_FAIL_ON_RETRY=1) since the rail settles; a press with
no pointerdown is a failure, not a retry.
- Assertion-gating sleeps 27 -> 1; the rejected-key "no requests" check
has a positive control.
Verified locally: strict fixture gate on the production build 28 ok + 2
"no Google 3D" skips (twice, fail-on-retry on); panel gate 5/5 plus no
render errors; live gate with a real token 25 ok + 3 fixture-only skips.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A mutation review left 18 earlier survivors; each now has a test that
fails with the mutant applied:
- the expanded viewer returns into the panel on shrink, close and destroy
(the fake DOM had no parentNode, so reinsertion was never exercised); the
fake DOM gains offsetParent and a MutationObserver that honours its
options, and a <body> child added while expanded goes inert;
- each pose moves the marker, drives FOLLOW and selects a new sequence;
close() retires an open still mounting or opening; a provider toggled
while the layer is off activates nothing; the adapter drops late and
superseded poses;
- share-link defaults: "any date" (0 days), the codec max pinned to the
filter's MAX_SINCE_DAYS (layerState cannot import it across package
boundaries, so a test ties the two), Mapillary on when a link omits it;
- Esc that something else handled, or typed in a field, keeps the
selection; a key problem outranks LOADING on the pill;
- the marker's terrain height (mesh, DEM, clamp then re-place);
- proxy: LRU memory cache and its half-budget rule, Retry-After capped,
a request after a cancelled fetch starts a fresh one;
- date-line tile ranking with and without focus; the horizon cull idles
while the camera stands still; the old zoom goes as soon as the new one
lands; a rate limit holds every refresh inside its wait.
One real bug found and fixed: Esc inside a contenteditable cleared the
selected sequence (selection.js now ignores isContentEditable targets).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
From the IC8 review of #768, each with tests that fail before:
- P2: the expanded viewer is a real modal: while it is on screen every
other <body> child is inert (only the ones it set are restored, on
shrink, close and destroy, and when Clean View/recording/cockpit hide
it). The voice pill and toasts are not clickable while it is expanded.
- P2: the tile proxy follows redirects by hand, at most 3 hops, each one
HTTPS on exactly https://tiles.mapillary.com without credentials;
anything else is a 502 and the token never leaves that origin.
- P2: the real dev and preview server smoke tests request
/api/mapillary/status and a tile, and the browser gate first checks the
server's real status route before any interception, so CI fails if the
Mapillary plugin is not registered.
- P2: both gates press panel headers through one shared helper
(scripts/qa-panelDrag.mjs): re-resolve the header until it is under the
pointer, record what the press hit, wait on the real floating/docked
state, and retry only a press that provably missed the header.
Verified: unit 5910 pass; fixture gate 17/17; panel gate 5/5; live gate
with a real Mapillary token 22/22.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
From gekh's review of #768, each with a test that fails before:
- The routes refuse cross-site requests (sameSiteGated, as the paid
proxies do) and the tile route has a generous per-IP limit (600/min), so
another page cannot burn the token's quota or trip the shared 429 hold.
- At most six upstream fetches run at once; a queued request that is
abandoned never fetches, nor does one abandoned during the disk read.
- A body past the 48 MB cap aborts the upstream request as it streams.
- Empty tiles are charged against the memory and disk budgets.
- Each disk write uses its own temp file and a retrim is skipped while a
write of that tile is pending, so concurrent hits cannot extend a tile's
life or fail a rename.
- Upstream refusals other than 401/403/429 answer 502.
- The tile route's success path is tested end to end (200 with the image
layer stripped and the right headers, 204 when empty, an abandoned
response aborts upstream); the teardown race that made this file flaky
is gone (240 parallel runs, no failure); the QA-constant tautologies are
replaced by a check against the real provider registry; the boundary
group lists the modules the provider actually imports.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
From the IC8 review of #768 (P2s), each with a test that fails before:
- A tile read from disk keeps its real age in memory, and the trim rewrite
keeps the file's mtime, so nothing is served past the 24 h TTL.
- The disk cache is swept in the background at most every 15 min: expired
tiles go, then the oldest past 1 GiB. Coverage zooms the app never
requests (z6-10) are refused with a 400, and layer names are matched as
own keys only.
- The tests run against a temporary cache directory (a test seam) instead
of writing into and deleting from the developer's real .gev-cache.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Tilted views: the visible box came from screen rays against the bare
ellipsoid and the 9 tiles kept were the ones nearest its centre, so a view
over high ground (Denver, 1,600 m up) or near the horizon dropped the
streets in front of the camera. At street zooms the rays now meet an
ellipsoid raised to the ground under the camera, hits beyond a range tied
to camera height are ignored, tilted views also keep the ground around the
camera, and tiles are ranked along the line of sight (nadir to screen
centre). Simulated with real Cesium ray maths over Sacramento and Denver
from 2 m to 1,000 m: the nearest visible ground and the screen centre are
always loaded.
Rejected token: Mapillary's 401/403 reached the panel as a raw "Mapillary
tiles HTTP 403" while the pill said ON, and every pan asked Mapillary again.
The proxy now answers a refusal as {keyRejected: true} and holds it for that
token (5 min), and a 429 with its Retry-After (held until then). The panel
reads KEY REJECTED, names MAPILLARY_CLIENT_TOKEN and Provider Settings,
gates its controls like KEY REQUIRED, requests nothing more, and clears the
message when the layer goes off. A rate limit keeps the drawn tiles and
refreshes once the wait is over.
Found in the IC8 review of #768 (P1 tile selection, P2 rejected token).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Fixes from gekh's re-review of 061d269, each with a regression test that
fails on the old code:
- A pan that drops a tile still loading now settles the loading count
(the aborted request no longer owns its key, so its `finally` never
did), and purges the previous zoom's tiles once nothing is pending.
LOADING no longer sticks on the status pill and chip.
- Zoom 0 (above 15,000 km) is a zoom: `zoom == null` instead of `!zoom`.
- A pose that arrives after a photo was closed is dropped, in the
MapillaryJS adapter and in the viewer host, so the marker and sequence
selection do not come back.
- Ground casting passes the tile's signal to the terrain request, so a
cancelled tile stops its height batch instead of holding the queue.
- The server's in-memory tile cache honours the 24 h TTL like the disk.
- The viewer's one-time resize runs for every photo, not just the first.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
preserve Bilawal's merge history, OpenAI schema-combinator startup fix, and Contacts viewport note
keep the validated voice and analyst behavior while adopting the current provider-neutral feed routes and Google token fallback
The Street Level and panel gates never ran in CI, and the live gate needs a
Mapillary token, the Mapillary network and real coverage at its test site,
so everything users see was guarded only by manual runs (which a bad token
and expired tile caches blocked twice this week).
`qa-street-level.mjs --fixtures` (npm run qa:street-level:fixtures) answers
the page's /api/mapillary/status and coverage-tile requests from generated
tiles (a street grid with flat and 360 deg sequences, half captured three
years ago; overview points from orbit) and blocks mapillary.com. Only the
six steps that open a photo are skipped. Because the data is known, the 360
deg and SINCE steps also assert that the filters remove sequences, not just
that counts do not grow. A unit test pins the fixture contents.
New CI job "Street Level browser gate": install with headless Chrome, start
a key-less dev server, run the fixture gate and qa:panel-resize. Locally,
against a dev server with every key blanked: 14/14 on three consecutive
runs, panel gate all pass.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Keeps the voice-core prompt rules and adds main's world-query and
vessels-vs-visuals rules; the tool catalog wraps the deixis-aware runner.
Also fixes two things the branch had before the merge:
- Live voice could not start: OpenAI rejects track_entity's top-level
anyOf. Model-facing tools now omit top-level combinators; the runner
and benchmark still require a query or a referent.
- The Contacts-centred analyst answer dropped the viewport coverage note.
The flight, military and vessel feeds were named after the providers the
bundled server uses (/api/opensky, /api/adsblol/mil, /api/ais-live and
their track routes; createOpenSkySource, createAdsbLolSource,
createAisStreamSource). Any server should be able to answer a feed from
its own provider, so the feeds are now:
/api/flights /api/flights/track?icao24=
/api/military /api/military/track?hex=
/api/vessels /api/vessels/track?mmsi=
with createFlightSource, createMilitarySource and createVesselSource. The
old routes are removed. Each server names its provider in X-Feed-Source
(the bundled server sends OpenSky Network, adsb.lol and AISStream, so its
attribution is unchanged) and the military feed's cache state and age in
X-Feed-Cache and X-Feed-Age-Ms. Sources report the provider the server
names and fall back to the feed's name.
Brings in 123 commits from main, including the v0.2.0 and v0.2.1 releases
(MCP server and tool catalog, server security hardening, build hygiene).
Conflicts:
- package.json: kept main's script order and added this PR's
qa:panel-resize and qa:street-level; exports keep both sides (main's
tools/view entries and this PR's street-level and Mapillary entries).
- src/tooling/viteBuild.test.mjs: kept this PR's Mapillary-token define
assertion and main's new wildcard-host allowedHosts assertions.
- CHANGELOG.md: main now has release sections; this PR's entries go under
[Unreleased].
Checks after the merge: doctor, format:check, check:boundaries,
layer-token:check (28 published, 1 new: street-level 0), npm test (5726
pass), build, the Windows job's focused tests (70 pass), qa:panel-resize,
and the keyless street-level gate. The new CSP allows Mapillary (https:
connect/img, blob workers). The keyed gate stops at coverage locally
because the dev .env Mapillary token is invalid (upstream 403), not because
of the merge.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The Content-Security-Policy blocked the YouTube player API, the Facebook
video SDK and X's post widget that src/data/bhoteKoshiEmbeddedMedia.js
loads, along with their players and posts in frames. script-src now allows
exactly those three script origins and a new frame-src allows YouTube,
youtube-nocookie, Facebook and X embeds. A test ties the policy to the
frames the media module builds and pins the allowed script origins.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
/api/google/text-search spends the same Google quota as nearby-places but
was not behind the cross-site gate. It now refuses cross-site requests
before calling Google, and a test covers both paid Google routes.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Vite installs its Host check after the middleware plugins add in
configureServer, so the provider and /api routes answered any Host,
including a DNS-rebinding name. hostCheckPlugin runs first on the dev and
preview servers and applies Vite's rule with the same allowed hosts: IP
addresses, localhost and *.localhost, the configured server host, and the
listed names. A test drives a real Vite server with an unlisted Host.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The voice debug log writes the server's timestamp after the client's record,
so a record can no longer replace loggedAt. The dev and preview servers also
send X-Content-Type-Options: nosniff with their shared browser headers.
From the findings in #710.
Co-authored-by: Sunil <[email protected]>
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Binding HOST=0.0.0.0 no longer sets Vite's allowedHosts to true. The dev and
preview servers accept IP addresses, localhost and the LAN hostnames listed
in GEV_ALLOWED_HOSTS; suffix and wildcard entries are ignored. The
configuration factory takes the resolved list as an input, so build/vite.js
still reads no environment.
Ported from #97 to the current build/vite.js layout. Fixes#21.
Co-authored-by: Sameh Khamis <[email protected]>
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
The cost-bearing endpoints (/api/realtime/token, /api/openai/hud-summary,
/api/google/nearby-places) and the voice debug log refuse cross-site browser
requests: a foreign or opaque Origin, a Sec-Fetch-Site other than
same-origin/none, or proxy forwarding headers. Loopback tools without an
Origin and the LAN opt-in keep working. The dev and preview servers send one
shared Content-Security-Policy, and embed-mode framing rewrites only the
frame-ancestors directive of it.
Salvaged from #242 onto current main.
Co-authored-by: Sameh Khamis <[email protected]>
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Embed-mode documents now keep the server's framing protection unless
GEV_EMBED_FRAME_ANCESTORS names the pages that may frame them (or * for any
page). The MCP Apps panel loads the app into its own page and needs no
framing. View answers are posted to the origin that sent the view rather
than to any origin.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Cesium's script sets CESIUM_WORKERS itself when it loads, so the panel's
combined workers script was never used, and neither was the embedded
terrain-heights file: the worker still requested it from the panel
page's site. The panel build now writes only the prelude, and the panel
puts it ahead of whatever Cesium assigns to CESIUM_WORKERS. The separate
esbuild bundle of Cesium's workers is gone.
The ground-polyline worker loads approximateTerrainHeights.json itself,
and a worker's requests go to the panel page's own site, where they
failed with 404 and stopped rendering once ground-clamped lines such as
military installations appeared. The panel build now embeds that file in
the workers script, which answers the worker's request from memory.
panel_request resolved paths with a check for a leading //, but URL
parsing also reads /\host as another host, so it could request other
servers. It now resolves the path against the app's address, refuses
any other origin, matches refused routes decoded as well, and does not
follow redirects. Responses are capped at 64 MiB while reading, with a
deadline; held responses make room on arrival and expire by time only,
so a response being read keeps its parts.
The panel keeps converted images only for its build files, so live
camera frames no longer accumulate, and failed build loads are retried.
The dev server answers a malformed /panel/ path with 400 instead of
crashing.
Hosts serve panels from their own sites and may refuse other addresses;
Codex refuses any address on the user's machine. The panel now never
requests the app's server: it loads the app's panel build and data
through panel_request, a tool only the panel may call, which requests
the path from the app's server. One path in every host; only map
providers load directly.
npm run build:panel writes the panel build to dist/panel, served at
/panel/: one app script, one stylesheet, Cesium's script, and Cesium's
workers as one script Cesium runs from memory. Tools may now be marked
app-only. The panel CORS rules are gone.
Codex serves panels from codex-sandbox:// pages, where origins are
opaque: the panel recognizes the app's own addresses by scheme and host,
the vessels feed resolves against the document address, and the dev
server answers Codex's sandbox origins with CORS.
A panel page loads the app with <base> pointing at the app's server, so
addresses the app builds resolve against the document base instead of
the page location, and hash updates stay on the page's own address.
The dev server answers panel origins with CORS on the API routes too.
show_on_globe replaces open_in_gods_eye_view and names an MCP Apps view, a
ui:// panel that loads the app in embed mode and posts each later view to it,
so the globe in the conversation changes without reloading. It also takes a
view another answer returned. Tools can declare a UI resource, and the MCP
adapter serves resources/list and resources/read and lists each tool's
resource in _meta.ui.resourceUri.
?embed=1 shows the globe alone: clean view with the HUD, panels, welcome and
setup prompts hidden, and provider attribution kept. The page that frames it
sends views with postMessage; the app applies each through its own actions
and reports every step. Annotations in any link are drawn once it restores.
Only embed-mode documents may be framed; every other page keeps
X-Frame-Options DENY and frame-ancestors 'none', and
GEV_EMBED_FRAME_ANCESTORS can restrict who frames them.
Voice hears audio directly, and "vessels" is easily heard as "visuals",
which the instructions route to visual styles, so asking for vessels turned
on other layers or a CRT style. The instructions now say a vessels request
without a style name means the ships layer, and a style needs its name.
The voice instructions recommended aircraft_in_area, vessels_in_area and
get_earthquakes, which voice no longer offers, and the docs said voice gets
nearly every query. The instructions now name tools voice has and send
aircraft, ship and hazard questions to analyst_query; a test keeps the
instructions to tools the voice session lists.
One add/add conflict, in scripts/format-scope.json: upstream's twelve new
entries and this branch's two. Both kept — 866 entries on main, 868 here,
none dropped.
Gates on the result: npm test 5397 passing, build, test:track 109/109,
changelog:check, format:check, check:boundaries, layer-token:check.
keep contacts counts, rows, scope and provenance on one retained snapshot
report capped cohorts as lower bounds
separate sibling tool cancellation from newer user turns
align track_entity validation across runtime and benchmark paths
