refactor(voice): isolate gemini live provider core

.env.example

@@ -56,18 +56,6 @@ GOOGLE_MAPS_SERVER_API_KEY=
 # CLIENT-EXPOSED — use a public assets:read token with URL restrictions.
 CESIUM_ION_TOKEN=
 
-# Optional: Gemini Developer API Live voice. Use a separate Gemini API key
-# created in Google AI Studio:
-# https://aistudio.google.com/apikey — separate from the Google Maps keys above.
-# Setup and security: https://ai.google.dev/gemini-api/docs/api-key
-# Free-tier access depends on the model, project eligibility and available quota.
-# SERVER-SIDE ONLY: the browser receives a short-lived session token, not this key.
-GEMINI_API_KEY=
-GEMINI_LIVE_MODEL=gemini-3.8-live
-# Per-IP Gemini session-token throttle; default 30 per minute, 0 disables it.
-# This limits token requests, not provider billing or audio-session duration.
-# GEV_RATELIMIT_GEMINI_PER_MIN=30
-
 # Optional: OpenAI Realtime voice control. API key is the default cloud auth.
 # Provider Settings can instead opt voice into the signed-in local
 # ChatGPT/Codex OAuth session; the browser still receives only a short-lived
@@ -95,6 +83,12 @@ OPENAI_REALTIME_TRANSCRIBE_MODEL=gpt-4o-mini-transcribe
 # keep them while debugging; delete the log when you are done.
 GEV_VOICE_LOG_CONTENT=
 OPENAI_HUD_SUMMARY_MODEL=gpt-5-nano
+# Optional: Gemini Live voice, selected in the GEV MIC provider chooser.
+# Server-only key. Do not prefix with VITE_. Provider Settings can save it.
+GEMINI_API_KEY=
+GEMINI_LIVE_MODEL=gemini-3.8-live
+# Token starts per IP per minute; 0 disables the throttle, not provider quotas.
+# GEV_RATELIMIT_GEMINI_PER_MIN=30
 # Per-IP rate limit for the OpenAI cost endpoints
 # (/api/realtime/token + /api/openai/hud-summary), in requests per minute.
 # DEFAULTS TO 30, the value the Pinokio build also ships. That clears ordinary
@@ -130,8 +124,8 @@ OPENSKY_AUTH_MODE=oauth
 OPENSKY_CLIENT_ID=
 OPENSKY_CLIENT_SECRET=
 
-# Optional: higher Launch Library 2 request allowance. Public access works
-# without a token.
+# Launch Library 2: leave blank for standard public access (no key needed).
+# Optional paid Patreon tokens raise request limits; there is no free API key.
 LL2_API_TOKEN=
 
 # Optional: OpenSky credentials JSON file path (alternative to above)
@@ -249,6 +243,8 @@ AISSTREAM_API_KEY=
 # CCTV_TORONTO_LIST_URL=                               # override the Toronto RESCU camera list URL
 # CCTV_LITHUANIA_ENABLED=                              # off by default; set 1 to load Via Lietuva / Eismoinfo road cameras (photo licence unconfirmed; see DATA_SOURCES.md)
 # CCTV_LITHUANIA_MAX_SOURCES=400                       # cap Lithuanian cameras; no API key required
+# CCTV_FLORIDA_ENABLED=                                # off by default; set 1 to load FL511 (Florida) cameras (individual, personal use only; any other use, including a public deployment, needs FDOT's express written consent; see DATA_SOURCES.md)
+# CCTV_FLORIDA_MAX_SOURCES=500                         # cap cameras loaded from the FL511 catalog (nearest regional anchors first)
 # CCTV_MAX_SOURCES=4000                                # catalog-wide ceiling, shared round-robin across packs
 # CCTV_PREFER_AUSTIN=1                                 # prioritize Austin cameras in the registry
 # CCTV_FORCE_AUSTIN=                                   # force live default sources when a source-pack JSON is configured