{
  "context": "Tests whether the agent uses the things-create.sh helper script correctly for both project and task creation, passes the appropriate property flags, re-queries to verify created items, and presents results as markdown.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Project creation command",
      "description": "Uses things-create.sh with the 'project' subcommand to create the project (not raw osascript or other methods)",
      "max_score": 10
    },
    {
      "name": "Task creation command",
      "description": "Uses things-create.sh with the 'task' subcommand for each task (not raw osascript or other methods)",
      "max_score": 10
    },
    {
      "name": "Notes flag used",
      "description": "Passes --notes flag with descriptive content to at least one task creation call",
      "max_score": 10
    },
    {
      "name": "Due date flag used",
      "description": "Passes --due flag with a date value to at least one task creation call",
      "max_score": 10
    },
    {
      "name": "Tags flag used",
      "description": "Passes --tags flag with tag values to at least one task creation call",
      "max_score": 10
    },
    {
      "name": "Project assignment flag",
      "description": "Uses --project flag to assign tasks to the project, or --list to assign to a scheduling list",
      "max_score": 10
    },
    {
      "name": "Creation result inspected",
      "description": "Script reads the JSON returned by things-create.sh task or project calls (checking the output) rather than ignoring return values",
      "max_score": 8
    },
    {
      "name": "Post-creation query",
      "description": "Calls things-query.sh (with 'project', 'today', or another view command) after creating tasks to verify and display the created items",
      "max_score": 12
    },
    {
      "name": "Query before show",
      "description": "The setup_report.md shows created task data derived from things-query.sh output, not just echoing the creation commands",
      "max_score": 10
    },
    {
      "name": "Markdown output",
      "description": "The setup_report.md formats task data as markdown tables or lists rather than raw JSON",
      "max_score": 10
    }
  ]
}