{
  "context": "This evaluation assesses how effectively the engineer uses Kedro's pipeline and node APIs to implement a pipeline failure recovery tool. The focus is on correctly utilizing Kedro's DAG construction, node access, and dependency resolution capabilities to identify resumption points using breadth-first search.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Uses node factory",
      "description": "Uses Kedro's node() factory function or Node class to create pipeline nodes with proper inputs and outputs specification",
      "max_score": 15
    },
    {
      "name": "Uses pipeline factory",
      "description": "Uses Kedro's pipeline() factory function or Pipeline class to construct the DAG from nodes",
      "max_score": 15
    },
    {
      "name": "Accesses pipeline nodes",
      "description": "Correctly accesses nodes from the Pipeline object using the .nodes property or similar API to iterate through the DAG",
      "max_score": 15
    },
    {
      "name": "Extracts node metadata",
      "description": "Properly accesses node attributes such as .name, .inputs, and .outputs to gather information about each node's dependencies and products",
      "max_score": 20
    },
    {
      "name": "Implements BFS logic",
      "description": "Implements breadth-first search algorithm to traverse the pipeline DAG and identify nodes that need to be re-run based on missing outputs",
      "max_score": 20
    },
    {
      "name": "Identifies resumption points",
      "description": "Correctly identifies the minimal set of resumption points (nodes to restart from) by analyzing which nodes have missing outputs and their dependencies",
      "max_score": 15
    }
  ]
}