Chat mode, phase 1: grounded streaming chat over your notes

Entry points (no sixth tab): Ask in the Library header (desktop button,
phone icon), Ask about this deck in the deck menu, Ask a question on
study/homework results (compact result travels as context), and a Chat
row in the desktop sidebar. Phones get a full-screen route with back +
New chat; the tab bar steps aside.

- Client-side retrieval: chunk scoped decks (~800 tokens), BM25 against
  the question + recent turns, top chunks under ~7k tokens, each
  labelled with deck + section. Notes come from the deck's stored
  sourceText (already kept for Regenerate); decks without it fall back
  to generated content and the UI says so.
- Stateless server: /api/chat validates against the CHAT_MODELS
  allowlist, trims the history to ~6k tokens, caps context at ~9k and
  streams token/thinking/done/error NDJSON. /api/chat/extract reduces
  attachments to text through the existing vision step.
- UI: user bubbles + full-width assistant text, streaming with stop,
  auto-scroll that pauses with a jump-to-bottom button, safe markdown
  (React nodes, no HTML injection; KaTeX dynamically imported only when
  a message has math), Thinking row for reasoning, copy/regenerate/
  edit-and-resend, model + context pickers (last model remembered),
  composer with visualViewport keyboard handling, safe-area, 16px
  textarea, Enter-to-send on desktop.
- Usage-limit failures keep the draft and show the shared message with
  Settings/Ollama actions; stopping maps the reader abort to a clean
  stopped state; offline disables sending with a reason.
- Settings: Keep notes text for chat toggle; deck menu: Delete notes
  text. RATE_LIMIT_CHAT + CHAT_MODELS in .env.example and README.
- Tests: chat suite (63 checks, stub Ollama) wired into run-all.
  Full set: 12/12 suites, 487 checks. SW -> freezmo-v32.

Also: tests/model-compare.mjs (diagnostic) benchmarks two models on the
same notes (latency, repair rounds, activity types) so the gpt-oss vs
gemma default can be decided on numbers; defaults unchanged.

.env.example

@@ -36,6 +36,13 @@ VISION_MODEL=gemma4:31b-cloud
 # Paid/credit: glm-5.3-flash:cloud, kimi-k3:cloud.
 TEXT_MODEL=gpt-oss:120b-cloud
 
+# CHAT MODELS (optional) -------------------------------------------------------
+# The allowlist the chat model picker offers, comma-separated. The server never
+# accepts a model name outside this list. The first entry is the default.
+# When unset, the configured TEXT_MODEL and VISION_MODEL are used.
+# Example: CHAT_MODELS=gemma4:31b-cloud,gpt-oss:120b-cloud
+# CHAT_MODELS=
+
 # RATE LIMITS (per client IP, sliding window, in-memory) ----------------------
 # Generation decks are expensive, so they are limited tightly by default.
 # RATE_LIMIT_WINDOW_MS=300000
@@ -49,6 +56,8 @@ TEXT_MODEL=gpt-oss:120b-cloud
 # RATE_LIMIT_SYNC=60
 # Push subscription changes and test sends.
 # RATE_LIMIT_PUSH=30
+# Chat turns (one model call each; users send them in bursts).
+# RATE_LIMIT_CHAT=60
 
 # PUSH REMINDERS (optional) ----------------------------------------------------
 # Web Push reminders only appear in the UI when BOTH VAPID keys are set.