feat: RSS embeddings, semantic news in chat, article-to-chat, richer briefings

- Embed RSS items at fetch time (nomic-embed-text); backfill at startup
- Semantic news search injected into chat system prompt ("Recent News You've Seen")
  when items match query above 0.55 cosine threshold (independent of note RAG)
- "Discuss in chat" button on news cards — creates a seeded conversation with
  the article title + full content, navigates directly to the new chat
- Briefing compilation now passes 500-char article excerpts (not just headlines)
  to the LLM and uses 8192 num_ctx to accommodate the larger prompt

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

src/fabledassistant/services/rss.py

@@ -112,14 +112,18 @@ async def fetch_and_cache_feed(feed_id: int, url: str) -> int:
         # only writes to items it successfully classifies, so already-classified items
         # are not re-processed (they have classified_at set).
         unclassified_ids: list[int] = []
+        new_item_data: list[tuple[int, str, str]] = []  # (id, title, content) for embedding
         if new_count > 0:
             result = await session.execute(
-                select(RssItem.id).where(
+                select(RssItem.id, RssItem.title, RssItem.content, RssItem.classified_at).where(
                     RssItem.feed_id == feed_id,
-                    RssItem.classified_at.is_(None),
                 )
             )
-            unclassified_ids = list(result.scalars().all())
+            for row in result.fetchall():
+                item_id, title, content, classified_at = row
+                if classified_at is None:
+                    unclassified_ids.append(item_id)
+                new_item_data.append((item_id, title or "", content or ""))
 
     # Prune old items to keep DB tidy
     await _prune_old_items(feed_id)
@@ -129,6 +133,17 @@ async def fetch_and_cache_feed(feed_id: int, url: str) -> int:
         from fabledassistant.services.rss_classifier import classify_and_store
         asyncio.create_task(classify_and_store(unclassified_ids, feed_user_id))
 
+    # Fire-and-forget embedding for new items
+    if new_item_data and feed_user_id is not None:
+        from fabledassistant.services.embeddings import upsert_rss_item_embedding
+
+        async def _embed_new_items() -> None:
+            for item_id, title, content in new_item_data:
+                await upsert_rss_item_embedding(item_id, feed_user_id, title, content)
+                await asyncio.sleep(0.05)
+
+        asyncio.create_task(_embed_new_items())
+
     return new_count