fix: rss classifier think-tag stripping, briefing calendar dict access, embed empty-string guard

- rss_classifier: strip <think>...</think> blocks (qwen3 reasoning output)
  before JSON parse; use strict=False for control chars; bump timeout 30s→120s
- briefing_pipeline: list_events returns dicts not Event objects — fix
  attribute access (.all_day/.start_dt/.title) to dict access
- embeddings: guard upsert_note_embedding and semantic_search_notes against
  empty-string input to prevent Ollama embed 400 errors

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

src/fabledassistant/services/rss_classifier.py

@@ -7,6 +7,7 @@ Called from rss.py after new items are stored — fire-and-forget.
 
 import json
 import logging
+import re
 from datetime import datetime, timezone
 
 import httpx
@@ -37,7 +38,7 @@ async def _llm_classify(prompt: str, model: str) -> str:
         "stream": False,
         "options": {"num_ctx": 2048, "temperature": 0.0},
     }
-    async with httpx.AsyncClient(timeout=30.0) as client:
+    async with httpx.AsyncClient(timeout=120.0) as client:
         resp = await client.post(f"{Config.OLLAMA_URL}/api/chat", json=payload)
         resp.raise_for_status()
         return resp.json().get("message", {}).get("content", "")
@@ -75,13 +76,17 @@ async def classify_items_batch(
 
     try:
         raw = await _llm_classify(prompt, model)
+        # Strip <think>...</think> blocks emitted by reasoning models (e.g. qwen3)
+        raw = re.sub(r"<think>.*?</think>", "", raw, flags=re.DOTALL)
         # Extract JSON from response (LLM may wrap it in markdown)
         raw = raw.strip()
         if raw.startswith("```"):
             raw = raw.split("```")[1]
             if raw.startswith("json"):
                 raw = raw[4:]
-        parsed = json.loads(raw)
+        raw = raw.strip()
+        # Allow control characters that may appear in LLM-generated JSON strings
+        parsed = json.loads(raw, strict=False)
         return {int(k): v for k, v in parsed.items() if isinstance(v, list)}
     except Exception:
         logger.warning("RSS classification failed", exc_info=True)