fix(hud): lead with the state tag, keep feedState in get_hud_caption

Review fixes for the HUD summary language change:

- The translated feed-state tag now leads the one-line summary and the
  fallback line, so the readout's ellipsis cannot hide it.
- get_hud_caption checks the reply's feedState, returns it and appends the
  state tag; a reply naming another state keeps only the place.
- The model gets the layer names the user sees (localizeLayerName).
- Japanese and Chinese summaries are capped at 20 characters; pt, zh-CN and
  zh-Hant get explicit language names.
- A partial view is narrated as incomplete coverage, not as not live.
- provenanceTag and hudSummaryMatchesProvenance move to layerSnapshot.js so
  the tools package can use them without browser globals.
- docs/CURRENT-STATE.md, changelog and comments updated; redundant tests
  trimmed.
This commit is contained in:
German Khokhlov committed 2026-10-30 13:06:53 +04:00
1 parent 54750a01e5
commit f967218a04
10 files changed
+232 -80

No files matched your search

+5 -2
View File
@@ -5255,8 +5255,11 @@ and unreachable upstream (502/504) separately from road geometry.
### AI HUD Summary (June 2026)
- HUD `SUMMARY` readout requests a five-word intelligence-style summary from `/api/openai/hud-summary` (model `OPENAI_HUD_SUMMARY_MODEL`, default `gpt-5-nano`, minimal reasoning).
- Input is the live basemap label context (place/street/nearby-place labels + enabled layers) — the model is instructed not to infer from coordinates.
- Output is sanitized to exactly five words; falls back to the deterministic telemetry summary on error/timeout (5s abort); typewriter animation on update.
- Input is the live basemap label context (place/street/nearby-place labels + enabled layers, the layer names as the user sees them) and `feedProvenance`; the model is instructed not to infer from coordinates.
- Language (October 2026): the HUD sends `locale`, which is part of the dedupe signature. The proxy keeps a shipped locale (English otherwise) and names the language in the instructions and at the start of the input; Japanese and Chinese get a 20-character phrase instead of five words.
- The reply is structured (`text.format` JSON schema): `{summary, feedState}`, `feedState` being the snapshot state or `none`. For a view that is not nominal, a missing or different `feedState` falls back to the deterministic line. An accepted summary is shown after the HUD's own translated state tag (`STALE LIVE FLIGHTS | …`), which leads so the one-line readout never clips it.
- Output is sanitized to five words (20 characters for Japanese and Chinese); falls back to the deterministic telemetry summary on error/timeout (5s abort); typewriter animation on update.
- `partial` is a feed state of its own, ranked between `fallback` and `nominal`, so one partial layer no longer reads as off and partial plus nominal reads partial (HUD, voice view state, `analyst_query`, `brief`, the agent world packet). Its narration note says coverage is incomplete rather than not live.
- A summary asked for before the camera starts moving never paints over the next view: `moveStart` marks the summary dirty and bumps its revision, as `moveEnd` does, aborts the request in flight and stops a reply still typing; neither a reply nor a fallback line (on an error, the 5 s timeout, the abort, or a failed context build) paints once the revision has moved. A dropped request forgets its signature and a timeout counts as a failure, so the next 15 s tick asks again even over the same place. An interrupted typewriter reply also forgets its signature; fully displayed answers remain reusable.
### Place-search providers