feat: robust metric extraction with confidence score and proof snippets

- fixed Year-Prefix Bug in MetricParser
- added metric_confidence and metric_proof_text to database
- added Entity-Check and Annual-Priority to LLM prompt
- improved UI: added confidence traffic light and mouse-over proof tooltip
- restored missing API endpoints (create, bulk, wiki-override)

company-explorer/reproduce_metric_bug.py

@@ -0,0 +1,29 @@
+import sys
+import os
+
+# Add backend to path
+sys.path.append(os.path.join(os.getcwd(), 'backend'))
+
+from lib.metric_parser import MetricParser
+
+def test_parser():
+    parser = MetricParser()
+    
+    test_cases = [
+        ("80 (2020)", 80.0),
+        ("80 (Stand 2022)", 80.0),
+        ("1.005 Mitarbeiter", 1005.0),
+        ("375,6 Mio. €", 375.6),
+        ("ca. 50.000 m²", 50000.0)
+    ]
+    
+    for text, expected in test_cases:
+        result = parser.extract_numeric(text)
+        print(f"Input: '{text}' -> Result: {result} (Expected: {expected})")
+        if result != expected:
+            print("  FAILED")
+        else:
+            print("  PASSED")
+
+if __name__ == "__main__":
+    test_parser()