feat: add in-browser local model support for iPhone via WebLLM

Enable Timmy to run directly on iPhone by loading a small LLM into
the browser via WebGPU (Safari 26+ / iOS 26+). No server connection
required — fully sovereign, fully offline.

New files:
- static/local_llm.js: WebLLM wrapper with model catalogue, WebGPU
  detection, streaming chat, and progress callbacks
- templates/mobile_local.html: Mobile-optimized UI with model
  selector, download progress, LOCAL/SERVER badge, and chat
- tests/dashboard/test_local_models.py: 31 tests covering routes,
  config, template UX, JS asset, and XSS prevention

Changes:
- config.py: browser_model_enabled, browser_model_id,
  browser_model_fallback settings
- routes/mobile.py: /mobile/local page, /mobile/local-models API
- base.html: LOCAL AI nav link

Supported models: SmolLM2-360M (~200MB), Qwen2.5-0.5B (~350MB),
SmolLM2-1.7B (~1GB), Llama-3.2-1B (~700MB). Falls back to
server-side Ollama when local model is unavailable.

https://claude.ai/code/session_01Cqkvr4sZbED7T3iDu1rwSD

src/config.py

@@ -90,6 +90,17 @@ class Settings(BaseSettings):
     work_orders_auto_execute: bool = False  # Master switch for auto-execution
     work_orders_auto_threshold: str = "low"  # Max priority that auto-executes: "low" | "medium" | "high" | "none"
 
+    # ── Browser Local Models (iPhone / WebGPU) ───────────────────────
+    # Enable in-browser LLM inference via WebLLM for offline iPhone use.
+    # When enabled, the mobile dashboard loads a small model directly
+    # in the browser — no server or Ollama required.
+    browser_model_enabled: bool = True
+    # WebLLM model ID — must be a pre-compiled MLC model.
+    # Recommended for iPhone: SmolLM2-360M (fast) or Qwen3-0.6B (smart).
+    browser_model_id: str = "SmolLM2-360M-Instruct-q4f16_1-MLC"
+    # Fallback to server when browser model is unavailable or too slow.
+    browser_model_fallback: bool = True
+
     # ── Scripture / Biblical Integration ──────────────────────────────
     # Enable the sovereign biblical text module.  When enabled, Timmy
     # loads the local ESV text corpus and runs meditation workflows.