src/feeds/rss_parser.py

@@ -6,6 +6,7 @@ import httpx
 from datetime import datetime, timezone
 from config import TIMEZONE, MAX_ARTICLES_PER_DOMAIN_RSS
 from source_rules import _extract_domain
+from feeds.transcript_extractors._common import html_to_text
 
 logger = logging.getLogger("osint.rss")
 
@@ -152,7 +153,11 @@ class RSSParser:
 
             for entry in feed.entries[:50]:
                 title = entry.get("title", "")
-                summary = entry.get("summary", "")
+                # RSS-summary ist bei vielen Quellen HTML (Guardian, AP, SZ, ...).
+                # Vor weiterer Verwendung strippen, sonst landet HTML in DB
+                # und KI-Agenten und Sprach-Heuristik werden gestoert.
+                summary_raw = entry.get("summary", "")
+                summary = html_to_text(summary_raw) if summary_raw else ""
                 text = f"{title} {summary}".lower()
 
                 # Adaptive Match-Schwelle: