{
  "context": "Tests whether the agent uses the batch estimation workflow with a consolidated summary table, estimates each task individually with per-task sizing, identifies cross-task dependencies, suggests an implementation order, and avoids redundant codebase analysis across tasks.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Consolidated summary table",
      "description": "Output includes a single table or matrix comparing all 5 features side-by-side with at least task name, time estimate, and size for each",
      "max_score": 12
    },
    {
      "name": "All five features estimated",
      "description": "Every one of the 5 features has an individual time estimate (not just a lump total)",
      "max_score": 10
    },
    {
      "name": "Per-task T-shirt size",
      "description": "Each feature is assigned its own T-shirt size (XS/S/M/L/XL)",
      "max_score": 8
    },
    {
      "name": "Grand total provided",
      "description": "Output includes a total time range summing across all 5 features",
      "max_score": 8
    },
    {
      "name": "Implementation order suggested",
      "description": "Output recommends a sequencing or priority order for the 5 features with reasoning",
      "max_score": 10
    },
    {
      "name": "Cross-task dependencies noted",
      "description": "Output identifies at least one dependency between features (e.g. roles needed before invitations, or activity feed depending on event model)",
      "max_score": 10
    },
    {
      "name": "Dark mode sized smallest",
      "description": "The dark mode feature receives the smallest estimate (XS or S) given strong existing support from Tailwind and shadcn/ui",
      "max_score": 8
    },
    {
      "name": "Permissions sized largest",
      "description": "The role-based permissions feature receives one of the larger estimates (M, L, or XL) reflecting its cross-cutting, security-critical nature",
      "max_score": 8
    },
    {
      "name": "Time ranges not points",
      "description": "All time values are expressed as ranges, not single point estimates",
      "max_score": 8
    },
    {
      "name": "Stack detected as TypeScript",
      "description": "Output identifies the stack as TypeScript/Next.js",
      "max_score": 5
    },
    {
      "name": "Per-task risk or confidence",
      "description": "Each feature has either a risk note or confidence level indicating relative uncertainty",
      "max_score": 5
    },
    {
      "name": "Shared assumptions listed",
      "description": "Output includes assumptions that apply across multiple or all features",
      "max_score": 8
    }
  ]
}