fix: always report broken internal links + show source pages

Previously, broken links that were part of the baseline ("known broken")
disappeared from the terminal output (score=0, no Klartext mention).

- differ.py: build reverse link index to record which pages link to
  each broken URL; "pages" field added to new_broken/known_broken entries
- plain.py: klartext_befunde() now includes known_broken_urls alongside
  new_broken_urls — broken links are always reported until fixed
- __main__.py: report shows "gefunden auf: <page>" for every broken entry,
  known entries labelled "[bekannt, noch nicht behoben]"
- 3 new tests (169 total)

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

scanner/__main__.py

@@ -299,8 +299,12 @@ def _render_markdown(r: dict) -> str:
         lines += ["", "## Kaputte Links", ""]
         for entry in new_broken:
             lines.append(f"- [NEU] HTTP {entry['status']} `{entry['url']}`")
+            for page in entry.get("pages", [])[:3]:
+                lines.append(f"  - gefunden auf: {page}")
         for entry in known_broken:
-            lines.append(f"- [bekannt] HTTP {entry['status']} `{entry['url']}`")
+            lines.append(f"- [bekannt, noch nicht behoben] HTTP {entry['status']} `{entry['url']}`")
+            for page in entry.get("pages", [])[:3]:
+                lines.append(f"  - gefunden auf: {page}")
 
     if d.get("unexpected_jsonld"):
         lines += ["", "## Unerwartetes JSON-LD", ""]

scanner/differ.py

@@ -216,12 +216,27 @@ def compare_snapshots(baseline_pages: dict, current_pages: dict, cfg: dict) -> d
         if p.get("status", 0) >= 400
     }
     current_broken = set(current_broken_map)
+
+    # Reverse index: broken_url → pages that contain a link to it
+    link_sources: dict[str, list[str]] = {}
+    for page_url, page in current_pages.items():
+        for link_list in page.get("links", {}).values():
+            if not isinstance(link_list, list):
+                continue
+            for link in link_list:
+                if isinstance(link, dict):
+                    href = link.get("url", "")
+                    if href:
+                        link_sources.setdefault(href, []).append(page_url)
+
     new_broken = [
-        {"url": url, "status": current_broken_map[url]}
+        {"url": url, "status": current_broken_map[url],
+         "pages": sorted(set(link_sources.get(url, [])))}
         for url in sorted(current_broken - baseline_broken)
     ]
     known_broken = [
-        {"url": url, "status": current_broken_map[url]}
+        {"url": url, "status": current_broken_map[url],
+         "pages": sorted(set(link_sources.get(url, [])))}
         for url in sorted(current_broken & baseline_broken)
     ]
 

scanner/plain.py

@@ -144,11 +144,14 @@ def klartext_befunde(
             "Nach einem geplanten Update ist das normal — sonst bitte prüfen lassen."
         )
 
-    # --- Kaputte Links (rein informativ) ---
+    # --- Kaputte Links (immer melden, bis behoben) ---
     neue_kaputt = diff.get("new_broken_urls", [])
-    if neue_kaputt:
+    bekannte_kaputt = diff.get("known_broken_urls", [])
+    alle_kaputt = len(neue_kaputt) + len(bekannte_kaputt)
+    if alle_kaputt:
         saetze.append(
-            f"{len(neue_kaputt)} interne Link(s) führen ins Leere (Seite nicht gefunden)."
+            f"{alle_kaputt} interne Link(s) führen ins Leere (Seite nicht gefunden). "
+            "Bitte prüfen und den Verweis auf der betroffenen Seite korrigieren."
         )
     if ext_links and ext_links.get("broken"):
         saetze.append(

tests/test_differ.py

@@ -234,6 +234,20 @@ class TestBrokenLinks:
         assert len(result["new_broken_urls"]) == 1
         assert result["new_broken_urls"][0]["url"] == "https://b.info/missing/"
         assert result["new_broken_urls"][0]["status"] == 404
+        assert "pages" in result["new_broken_urls"][0]
+
+    def test_broken_url_includes_source_pages(self):
+        """broken entry lists the pages that link to the broken URL."""
+        home = self._make_page("https://b.info/")
+        home["links"]["a"] = [{"url": "https://b.info/missing/", "class": "internal"}]
+        baseline = {"https://b.info/": self._make_page("https://b.info/")}
+        current = {
+            "https://b.info/": home,
+            "https://b.info/missing/": self._make_page("https://b.info/missing/", status=404),
+        }
+        result = compare_snapshots(baseline, current, BASE_CFG)
+        entry = result["new_broken_urls"][0]
+        assert "https://b.info/" in entry["pages"]
 
     def test_known_broken_not_in_new(self):
         broken_page = {"https://b.info/old/": self._make_page("https://b.info/old/", status=404)}
@@ -242,6 +256,7 @@ class TestBrokenLinks:
         result = compare_snapshots(baseline, current, BASE_CFG)
         assert result["new_broken_urls"] == []
         assert len(result["known_broken_urls"]) == 1
+        assert "pages" in result["known_broken_urls"][0]
 
     def test_broken_url_scores(self):
         result = score_diff({

tests/test_plain.py

@@ -70,6 +70,20 @@ class TestKlartextBefunde:
         assert any("funktionieren" in s.lower() or "website" in s.lower() for s in saetze)
         _no_jargon(saetze)
 
+    def test_new_broken_internal_link_plain(self):
+        diff = {"new_broken_urls": [{"url": "https://x.de/gone", "status": 404, "pages": []}]}
+        saetze = klartext_befunde(diff)
+        assert any("leere" in s.lower() or "gefunden" in s.lower() for s in saetze)
+        _no_jargon(saetze)
+
+    def test_known_broken_link_still_reported(self):
+        """known_broken_urls must appear in Klartext even with score=0."""
+        diff = {"known_broken_urls": [{"url": "https://x.de/old", "status": 404, "pages": []}]}
+        saetze = klartext_befunde(diff)
+        assert any("leere" in s.lower() or "gefunden" in s.lower() for s in saetze), \
+            "known_broken_urls müssen auch bei Score=0 im Klartext erscheinen"
+        _no_jargon(saetze)
+
 
 class TestWasTun:
     def test_red_has_urgent_steps(self):