feat(memory): fix(rollover): v1 extractor skips files at exactly max_lines

Co-Authored-By: @memory <memory@aipass>

src/aipass/memory/apps/handlers/rollover/extractor.py

@@ -387,7 +387,7 @@ def extract_items(file_path: Path, percentage: int | None = None) -> Dict[str, A
     max_lines = data.get("document_metadata", {}).get("limits", {}).get("max_lines", 600)
 
     # Check if under limit
-    if current_lines <= max_lines:
+    if current_lines < max_lines:
         return {"success": True, "skipped": True, "message": f"File under limit ({current_lines}/{max_lines} lines)"}
 
     # Calculate extraction amount (simulate actual line reduction)

src/aipass/memory/tests/test_rollover_pipeline.py

@@ -617,6 +617,41 @@ class TestExtractWithMetadata:
             assert "branch" in entry["_metadata"]
             assert "extracted_at" in entry["_metadata"]
 
+    def test_v1_extracts_at_exactly_max_lines(self, monkeypatch, tmp_path):
+        """v1 file at exactly max_lines should extract, not skip."""
+        ext, mocks = _import_extractor(monkeypatch)
+
+        observations = [
+            {"date": f"2026-01-{i:02d}", "session": i, "entries": [{"title": f"obs {i}"}]} for i in range(1, 11)
+        ]
+        data = {
+            "document_metadata": {
+                "schema_version": "1.0.0",
+                "limits": {"max_lines": 50},
+                "status": {},
+            },
+            "observations": observations,
+        }
+        file_path = tmp_path / "DEVPULSE.observations.json"
+        content = json.dumps(data, indent=2)
+        file_path.write_text(content, encoding="utf-8")
+        actual_lines = len(content.splitlines())
+
+        data["document_metadata"]["limits"]["max_lines"] = actual_lines
+
+        file_path.write_text(json.dumps(data, indent=2), encoding="utf-8")
+        mocks["memory_files"].read_memory_file_data.return_value = data
+
+        def fake_write(fp, d):
+            fp.write_text(json.dumps(d, indent=2), encoding="utf-8")
+
+        monkeypatch.setattr(ext, "_write_memory_file", fake_write)
+
+        result = ext.extract_items(file_path)
+        assert result["success"] is True
+        assert result.get("skipped") is not True
+        assert result["extracted_count"] > 0
+
     def test_skipped_result_passes_through(self, monkeypatch, tmp_path):
         """When extract_items returns skipped (under limit), extract_with_metadata passes it."""
         ext, mocks = _import_extractor(monkeypatch)