Merge main into feat/voice-core

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.

.env.example

@@ -34,16 +34,19 @@ GOOGLE_MAPS_API_KEY=
 # single shared key keeps working — splitting is opt-in.
 # SERVER-SIDE ONLY.
 GOOGLE_MAPS_SERVER_API_KEY=
-# Optional: opt-in per-IP rate limit for the Google Places cost endpoint
-# (/api/google/nearby-places), in requests per minute.
-# DEFAULT IS UNLIMITED — unset (or 0) means no throttling, unchanged behavior.
-# Set a positive integer to cap per-IP Places quota spend when self-hosting.
-# Recommended whenever HOST is not localhost. NOTE: this is a per-IP,
-# process-local guard (in-memory, resets on restart; over-limit requests get a
-# sanitized 429) — it is NOT a billing cap. For hard spend protection set
-# provider-side budget alerts: Google Cloud Console -> Billing -> Budgets & alerts
-# (plus per-API quotas under APIs & Services -> Quotas).
-# GEV_RATELIMIT_GOOGLE_PER_MIN=60
+# Per-IP rate limit for the Google Places cost endpoints
+# (/api/google/nearby-places + /api/google/text-search), in requests per minute.
+# DEFAULTS TO 120, the value the Pinokio build already ships. Raise or lower it
+# with a positive integer; 60 is a sensible tightening whenever HOST is not
+# localhost. Set it to exactly 0 to run unthrottled. An unreadable value (a
+# typo, a stray unit) falls back to the default rather than to unlimited, so a
+# slip cannot quietly disarm the guard. A positive value below 1 counts as 1.
+# NOTE: this is a per-IP, process-local guard (in-memory, resets on restart;
+# over-limit requests get a sanitized 429) — it is NOT a billing cap. For hard
+# spend protection set provider-side budget alerts: Google Cloud Console ->
+# Billing -> Budgets & alerts (plus per-API quotas under APIs & Services ->
+# Quotas).
+# GEV_RATELIMIT_GOOGLE_PER_MIN=60   # uncomment to tighten from the default 120
 
 # Optional: Google Photorealistic 3D Tiles through Cesium ion, Bing world
 # imagery, and Cesium World Terrain. The free Community plan is for eligible
@@ -72,15 +75,19 @@ OPENAI_REALTIME_TRANSCRIBE_MODEL=gpt-4o-mini-transcribe
 # Set to 1 to keep them while debugging; delete the log when you are done.
 GEV_VOICE_LOG_CONTENT=
 OPENAI_HUD_SUMMARY_MODEL=gpt-5-nano
-# Optional: opt-in per-IP rate limit for the OpenAI cost endpoints
+# Per-IP rate limit for the OpenAI cost endpoints
 # (/api/realtime/token + /api/openai/hud-summary), in requests per minute.
-# DEFAULT IS UNLIMITED — unset (or 0) means no throttling, unchanged behavior.
-# Set a positive integer to cap per-IP spend when self-hosting on a network.
-# Recommended whenever HOST is not localhost. NOTE: this is a per-IP,
-# process-local guard (in-memory, resets on restart; over-limit requests get a
-# sanitized 429) — it is NOT a billing cap. For hard spend protection set a
-# provider-side budget: OpenAI platform -> Settings -> Limits (usage limits).
-# GEV_RATELIMIT_OPENAI_PER_MIN=30
+# DEFAULTS TO 30, the value the Pinokio build also ships. That clears ordinary
+# use several times over: an open tab asks for a HUD summary every 15s (4/min)
+# and a voice session mints one token. Raise or lower it with a positive
+# integer; set it to exactly 0 to run unthrottled. An unreadable value (a typo,
+# a stray unit) falls back to the default rather than to unlimited, so a slip
+# cannot quietly disarm the guard. A positive value below 1 counts as 1.
+# NOTE: this is a per-IP, process-local guard (in-memory, resets on restart;
+# over-limit requests get a sanitized 429) — it is NOT a billing cap. For hard
+# spend protection set a provider-side budget: OpenAI platform -> Settings ->
+# Limits (usage limits).
+# GEV_RATELIMIT_OPENAI_PER_MIN=30   # the default; uncomment only to change it
 
 # Optional: OpenSky auth mode (oauth | basic | auto | anon)
 # Default: oauth — requires client credentials below
@@ -108,6 +115,11 @@ PORT=4173
 # networks you trust, and consider the GEV_RATELIMIT_* throttles above.
 # HOST=localhost
 # HOST=0.0.0.0
+# Optional: comma-separated LAN hostnames to accept, in addition to localhost
+# and IP addresses, for example GEV_ALLOWED_HOSTS=globe.lan,globe.internal.
+# Binding HOST=0.0.0.0 alone does NOT trust every Host header; suffix and
+# wildcard entries are ignored.
+# GEV_ALLOWED_HOSTS=
 
 # Optional: NASA FIRMS live active fires. SERVER-SIDE ONLY (never reaches the
 # browser — the client polls the /api/firms proxy). Free key:
@@ -193,3 +205,8 @@ AISSTREAM_API_KEY=
 
 # Optional: comma-separated Overpass URLs; use only instances you run or pay for.
 OVERPASS_UPSTREAMS=
+
+# Optional: who may frame the app in embed mode (?embed=1), as CSP
+# frame-ancestors sources such as "https://example.com", or * for any page.
+# Unset, no page may frame it. Other pages of the app can never be framed.
+# GEV_EMBED_FRAME_ANCESTORS=