{
  "context": "Evaluates whether the CLI helper actually delegates execution to Vedro while wiring its slicing and ordering controls so workers can run deterministic shards. Scoring prioritizes correct use of Vedro's CLI/config flags over custom scheduling logic.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Vedro entrypoint",
      "description": "Uses Vedro's runner (`vedro run` shell command or `vedro.run()` entrypoint) to execute scenarios instead of implementing a bespoke test loop.",
      "max_score": 25
    },
    {
      "name": "Slice forwarding",
      "description": "Passes the user-provided slice string directly into Vedro via `--slice <index>/<total>` or `Config.Plugins.Slicer.slice`, ensuring only that shard executes.",
      "max_score": 25
    },
    {
      "name": "Ordering modes",
      "description": "Maps CLI flags to Vedro's orderer options (`--order declaration|reversed|random` or `Config.Plugins.Orderer.orderer`) so the requested ordering mode reaches the framework.",
      "max_score": 20
    },
    {
      "name": "Random seed",
      "description": "When random ordering is chosen, forwards the seed through Vedro (`--seed <value>` or `Config.Plugins.Seeder.seed`) to keep slice order reproducible across runs.",
      "max_score": 20
    },
    {
      "name": "Arg passthrough",
      "description": "Extras after `--` are preserved when invoking Vedro (e.g., `vedro run ... -- <args>`), guaranteeing downstream options reach the test runner unchanged.",
      "max_score": 10
    }
  ]
}