{
  "context": "Evaluates whether the solution leverages standard library type stubs to enforce precise typing for filesystem paths, date/time scheduling, immutable tags, and iterable job collections in the scheduler. Checks reward correct annotations and usage of typed stdlib constructs rather than loosely typed values.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Path typing",
      "description": "Job loading accepts a pathlib.Path argument and performs file access with Path-aware types instead of raw strings.",
      "max_score": 15
    },
    {
      "name": "Datetime tz-aware",
      "description": "Start times and scheduled run times use timezone-aware datetime objects, preserving tzinfo from ISO inputs when constructing RunPlan entries.",
      "max_score": 20
    },
    {
      "name": "Timedelta intervals",
      "description": "Interval math relies on datetime.timedelta objects derived from minute values to advance schedules, not integer arithmetic.",
      "max_score": 15
    },
    {
      "name": "Command typing",
      "description": "Commands are represented as Sequence[str] and maintained as typed string sequences throughout parsing and planning.",
      "max_score": 15
    },
    {
      "name": "Tags immutability",
      "description": "Optional tags are stored as frozenset[str] | None and tag filtering operates using those immutable sets.",
      "max_score": 15
    },
    {
      "name": "Iterable handling",
      "description": "next_runs accepts Iterable[JobDefinition], returns a list[RunPlan], and orders results deterministically using typed comparisons.",
      "max_score": 10
    },
    {
      "name": "No loose typing",
      "description": "All public signatures and key intermediates avoid Any, instead using stdlib typing constructs (Optional, list, frozenset, Sequence) consistent with the spec annotations.",
      "max_score": 10
    }
  ]
}