feat: possession presentation layer — glitchy text + degraded audio

First sub-project of the "make contact feel real" arc (spec:
docs/superpowers/specs/2026-07-23-possession-presentation-design.md).
Direct Contact replies now feel like a spirit fighting through static to
hold the channel rather than a plain chat bubble:

- backend/app/possession.py: compute_stability(rarity, magnitude, rng) —
  a 0.05-0.98 score per reply (rarer entity + stronger triggering anomaly =
  cleaner signal), rng injectable for a later quantum-RNG source.
- ws.py sends stability on reply_start; audio synthesis for that reply gets
  noise/bitcrush scaled by instability (1 - stability) via a new
  instability param on synthesize_spirit_voice — effects.py itself is
  untouched, only the params fed into it.
- frontend/src/lib/possession.ts: renderPossessedText — pure, deterministic
  (tick-seeded, no Math.random) text corruption with self-correcting
  glitch bursts, wired into Transcript.tsx's streaming reply display.

Stored transcript/reply text is unaffected — this is presentation only.
78/78 backend, 137/137 frontend tests passing.

backend/tests/test_ws_session.py

@@ -31,7 +31,7 @@ class FakeSpiritService:
         return False
 
 
-async def _fake_synth(text, voice, profile):
+async def _fake_synth(text, voice, profile, instability=0.0):
     return b"RIFFfake wav bytes"
 
 
@@ -124,7 +124,8 @@ async def test_question_streams_reply_and_records_history(sync_client, db_sessio
         _read_until(ws, "session")
         ws.send_json({"type": "question", "text": "Are you at peace?"})
         _read_until(ws, "entity")  # auto-summoned before answering
-        _read_until(ws, "reply_start")
+        reply_start = _read_until(ws, "reply_start")
+        assert 0.05 <= reply_start["stability"] <= 0.98
         reply_end = _read_until(ws, "reply_end")
         assert reply_end["text"] == "I am here."