{
  "context": "Verify that the agent correctly identifies the high-cardinality risk, explains when the pattern is acceptable, and provides both a filtered alternative and a recording rule.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "High-cardinality risk is identified and explained",
      "description": "The report states that the query has no label filters and explains that this causes fan-out across all series for the metric (referencing the `path`, `status_code`, and multi-service context).",
      "max_score": 25
    },
    {
      "name": "Nuanced acceptability analysis provided",
      "description": "The report explains that `sum()` aggregation produces a single scalar result, so the query is acceptable IF the cluster can handle the fan-out — it does not simply label the query as 'bad'.",
      "max_score": 20
    },
    {
      "name": "Filtered per-service alternative is provided",
      "description": "The report includes a query that adds a `job=` or `service=` label filter to restrict the query to a single service.",
      "max_score": 20
    },
    {
      "name": "Recording rule definition is provided",
      "description": "The report includes a YAML recording rule definition that pre-aggregates the total request rate, reducing per-query scan cost.",
      "max_score": 20
    },
    {
      "name": "Semantic and syntactic correctness confirmed",
      "description": "The report confirms that the query is syntactically valid and semantically correct for computing total request rate (counter metric with `rate()` and `sum()`).",
      "max_score": 15
    }
  ]
}