{
  "context": "This criteria evaluates how effectively the engineer uses Kedro's unified filter interface (Pipeline.filter() method) and related pipeline filtering capabilities to implement a flexible pipeline filtering utility.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Uses Pipeline.filter()",
      "description": "Implementation uses the Pipeline.filter() method as the primary filtering mechanism, demonstrating knowledge of Kedro's unified filter interface",
      "max_score": 40
    },
    {
      "name": "Tag filtering",
      "description": "Correctly implements tag-based filtering using the tags parameter of Pipeline.filter() or Pipeline.only_nodes_with_tags()",
      "max_score": 15
    },
    {
      "name": "From-nodes filtering",
      "description": "Correctly implements from_nodes filtering using the from_nodes parameter of Pipeline.filter() or Pipeline.from_nodes()",
      "max_score": 10
    },
    {
      "name": "To-outputs filtering",
      "description": "Correctly implements to_outputs filtering using the to_outputs parameter of Pipeline.filter() or Pipeline.to_outputs()",
      "max_score": 10
    },
    {
      "name": "Combined filter logic",
      "description": "Correctly handles multiple filter criteria combined with AND logic, as specified by Kedro's filter() method behavior",
      "max_score": 15
    },
    {
      "name": "Return type",
      "description": "Returns a Pipeline object (not a list of nodes or other type), maintaining Kedro's pipeline abstraction",
      "max_score": 10
    }
  ]
}