feat: spirit engine — seance WS, entity minting/Codex, Piper TTS voices, wire telemetry

- WS /ws/session: modes, summon, anomaly fragments, streaming direct contact,
  passive wire-ghost ambient loop, per-user rate limits, event transcript
- entities: anomaly-signature fingerprinting, LLM minting + procedural
  fallback, Codex matching with contact counts and sightings
- tts: 8 local Piper voices (EN/ES), per-entity voice profiles, numpy
  effects chain (pitch/rate/bitcrush/echo/static)
- llm: streaming client, submit_stream in bounded queue, SpiritService
  with offline fallbacks for every channel
- routes: public /api/codex, /api/codex/{id}, /api/stats; /audio static mount
- models: Entity, EntitySighting, Event, ContactSession(entity_id, language)

backend/app/config.py

@@ -9,5 +9,18 @@ class Settings(BaseSettings):
     session_secret: str
     port: int = 7777
 
+    # LLM tiers — CPU-only remote Ollama, so the fast tier must stay small.
+    ollama_fast_model: str = "granite4.1:3b"
+    ollama_chat_model: str = "minicpm-v4.5:latest"
+    llm_max_concurrency: int = 2
+    llm_max_queue_depth: int = 8
+    llm_cooldown_seconds: float = 8.0  # min gap between ambient whispers
+
+    # TTS
+    piper_voices_dir: str = "voices"
+
+    # Where synthesized utterance audio is written (served at /audio/).
+    data_dir: str = "data"
+
 
 settings = Settings()