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/app/possession.py

@@ -0,0 +1,39 @@
+"""Possession stability: a single number expressing how cleanly a spirit
+holds the channel for one reply. Shared contract between the backend audio
+degradation and the frontend text-glitch renderer (see
+docs/superpowers/specs/2026-07-23-possession-presentation-design.md)."""
+
+import random
+from typing import Callable
+
+_BASE_BY_RARITY = {
+    "common": 0.35,
+    "uncommon": 0.5,
+    "rare": 0.65,
+    "mythic": 0.8,
+}
+
+_MIN_STABILITY = 0.05
+_MAX_STABILITY = 0.98
+_MAX_MAGNITUDE_BONUS = 0.15
+_JITTER_SPAN = 0.1  # rng() in [0, 1) is scaled to +/- this amount
+
+
+def compute_stability(
+    rarity: str,
+    magnitude: float,
+    rng: Callable[[], float] = random.random,
+) -> float:
+    """How cleanly the spirit holds the channel for this reply, 0.05-0.98.
+
+    Rarer spirits hold the channel more steadily (higher base). A stronger
+    triggering anomaly gives a cleaner line, up to a capped bonus. A final
+    rng-driven jitter of +/- 0.1 keeps it from feeling deterministic to the
+    seeker. `rng` is injectable so a later quantum-RNG source can swap in
+    real entropy without touching call sites.
+    """
+    base = _BASE_BY_RARITY.get(rarity, _BASE_BY_RARITY["common"])
+    magnitude_bonus = min(_MAX_MAGNITUDE_BONUS, magnitude / 100)
+    jitter = (rng() * 2 - 1) * _JITTER_SPAN
+    stability = base + magnitude_bonus + jitter
+    return max(_MIN_STABILITY, min(_MAX_STABILITY, stability))