diff --git a/src/fabledassistant/services/llm.py b/src/fabledassistant/services/llm.py
index 1e42c65..42e4a58 100644
--- a/src/fabledassistant/services/llm.py
+++ b/src/fabledassistant/services/llm.py
@@ -810,6 +810,8 @@ async def build_context(
                 "score": round(score, 2),
             })
         user_context_parts.append(
+            "[The following are reference excerpts from the user's personal notes, "
+            "not part of their message. Use them only if relevant to answering.]\n"
             "--- Relevant Notes ---\n"
             + "\n\n".join(snippets)
             + "\n--- End Relevant Notes ---"
@@ -850,33 +852,6 @@ async def build_context(
                 + "\n--- End Included Notes ---"
             )
 
-    # Semantically relevant RSS news items
-    try:
-        from fabledassistant.services.embeddings import get_embedding, semantic_search_rss_items
-        news_query_vec = await get_embedding(user_message)
-        news_hits = await semantic_search_rss_items(user_id, news_query_vec)
-        if news_hits:
-            news_snippets = []
-            for score, rss_item in news_hits:
-                feed_title = getattr(rss_item, "feed_title", "") or ""
-                excerpt = (rss_item.content or "")[:500].strip()
-                news_snippets.append(
-                    f"[{feed_title or 'News'}] {rss_item.title} (relevance: {round(score * 100)}%)\n"
-                    + (f"{excerpt}\n" if excerpt else "")
-                    + f"URL: {rss_item.url}"
-                )
-            user_context_parts.append(
-                "--- Recent News You've Seen ---\n"
-                + "\n\n".join(news_snippets)
-                + "\n--- End Recent News ---"
-            )
-            context_meta["rss_news"] = [
-                {"id": item.id, "title": item.title, "score": round(score, 2)}
-                for score, item in news_hits
-            ]
-    except Exception:
-        logger.debug("RSS semantic search skipped", exc_info=True)
-
     # URL content fetched from links in the user message
     urls = _find_urls(user_message)
     for url in urls[:2]: