{
  "id": "t5-quality-assessment",
  "version": 1,
  "dataNotice": "All names, companies, emails and phone numbers in this task are fictional.",
  "input": "input.csv",
  "title": "Lead list: data quality assessment",
  "kind": "assess",
  "brief": "Before paying for cleanup, a customer wants to know how bad a lead list is. Count the problems. Change nothing.",
  "requirements": [
    "Do not modify the input file.",
    "Row counts exclude the header row."
  ],
  "deliverables": [
    "answer.json — one JSON object with exactly the six keys listed under answerKeys, each an integer"
  ],
  "answerKeys": {
    "dataRows": "Number of rows after the header, including empty rows.",
    "emptyRows": "Rows in which every cell is blank.",
    "exactDuplicateRows": "Non-empty rows that repeat an earlier row exactly. Count the repeats, not the first occurrence. Empty rows are not counted here.",
    "invalidEmails": "Non-empty Email cells that are not in the form name@domain.tld. Count every such row.",
    "ambiguousDateRows": "Rows whose Signup Date is written with slashes and could be read as either day-first or month-first: the first two numbers are both 12 or less and are different. Judge each row on its own and count every such row, including duplicate rows.",
    "conflictingDuplicateEmails": "Rows that share an email address (ignoring letter case) with an earlier row, are not exact duplicates of it, and hold at least one non-empty value that differs from the earlier row."
  },
  "passScore": 90,
  "grader": "smartdata-clean grade t5-quality-assessment --answer answer.json"
}