feat(classify): super-linear positive weighting by hit count

正向改为按命中次数 n（require_title + reaction + charge + weak）做平方加权，使多信号同现的真 reaction 不被单个负向词打穿（1+1>2、1+1+1>5）；标题不含 require_title 直接 0 分。库内 796 条有效视频复测：线以下误伤由 47 降至 14。

tests/test_discovery.py

@@ -47,7 +47,7 @@ def write_config(tmp, **discover):
             "date_negative": True,
             "duration_boost": {"gt_30min": 1, "gt_60min": 2},
         },
-        "notify": {"min_score": 3, "max_items_per_message": 10},
+        "notify": {"min_score": 4, "max_items_per_message": 10},
     }
     path = Path(tmp) / "config.json"
     path.write_text(json.dumps(data, ensure_ascii=False), encoding="utf-8")
@@ -130,7 +130,7 @@ def test_filters_and_scoring():
             assert summary.blacklisted == 1
             assert summary.candidates == 1  # aid 6
             assert summary.notified == 0
-            assert repo.get_candidate(6)["score"] == 0
+            assert repo.get_candidate(6)["score"] == 1
             assert repo.get_candidate(2) is None
             assert repo.get_candidate(4) is None
             assert repo.get_candidate(5) is None