{
  "kind": "synthetic-acceptance-worksheet",
  "solution": "stockvexel",
  "title": "Three locations, three different decisions",
  "notice": "This invented reconciliation worksheet shows why a scan result is not an automatic stock adjustment. It is not output from a StockVexel run.",
  "columns": [
    "Location",
    "Expected pallet in the WMS snapshot",
    "Candidate observation",
    "Expected next step"
  ],
  "rows": [
    ["A-01", "P-101", "P-101", "Identifier matches; retain the observation"],
    ["A-02", "P-102", "P-109", "Review the location conflict and recent movements"],
    ["A-03", "P-103", "Unreadable label", "Recheck the label; do not record zero stock"]
  ],
  "interpretation": [
    "The worksheet gives one identifier match and two exceptions. It does not prove that a pallet is missing or that item quantities inside a pallet are correct.",
    "Before resolving A-02, compare the scan time with the WMS snapshot and any intervening receipt, pick or move. Before resolving A-03, obtain a reliable read or an independent physical check. An unreadable label is not a zero-stock observation.",
    "The measurement unit here is a uniquely identified pallet at a mapped location. A barcode read alone cannot verify its contents or quantities.",
    "Agree whether the integration reads inventory, creates review tasks or writes approved adjustments. Any approved adjustment should be traceable to the evidence and review that authorized it. These are proposed acceptance decisions, not tested integration behavior."
  ]
}