{
  "output": "# CSV quality report\n\nAssumption: the first CSV record is the header. Delimiter: comma; UTF-8. Quoted newlines stay inside a record.\nData record indexes below start at 1 after the header; they are not physical line numbers.\n\n- Records: 4 including header; data records: 3; header columns: 3.\n- Present data cells: 9; empty/whitespace cells: 1; absent header-aligned cells: 0; extra cells beyond header: 0.\n- Width mismatches: 0; first record indexes: none.\n- Completely blank data records: 0; first indexes: none.\n- Exact duplicate data records beyond their first occurrence: 1; first indexes: 3.\n- Empty header labels: 0; first column indexes: none.\n- Duplicate header groups after trimming (case-sensitive): 0.\n\n## Per-column checks\n\nMissing = an absent position or empty/whitespace text. NA/null words are not treated as missing. Numeric = decimal/exponent syntax; no values are coerced or inferred as dates.\n\n| # | Header (trimmed display, max 16 characters) | Missing | Numeric nonmissing | Other nonmissing | First missing record indexes |\n| --- | --- | ---: | ---: | ---: | --- |\n| 1 | id | 0 | 0 | 3 | none |\n| 2 | quantity | 1 | 2 | 0 | 2 |\n| 3 | note | 0 | 0 | 3 | none |\n\nIndexes are bounded to the first 8 examples per check. Duplicate records compare exact parsed field text, including width and whitespace. No input data is fetched, executed, corrected, stored, or sent to another service.\n"
}