Profile every column, author business definitions and ownership, mark sensitivity, validate the complete file, and reuse the reviewed contract on the next delivery — without uploading data.
A schema says what shape a field has. A data dictionary explains what it means, who owns it, how it may be used, and whether the next file still follows the reviewed contract.
Inspect type, completeness, distinctness, key/category candidates, and header-only personal-data signals locally.
Write dataset and field definitions, ownership, roles, types, sensitivity, units, formats, allowed values, and notes.
Check every row, expose schema drift, save a data-free definition, and export CSV, Markdown, HTML, JSON Schema, or JSON.
Authored meaning stays beside deterministic observed evidence, so a future delivery can be checked without turning the dictionary into a copy of the data.
Capture business names, definitions, owner, version, cadence, roles, units, formats, allowed values, and stewardship notes.
Header-only signals prompt review; personal-data and sensitivity classifications remain explicit, editable, and never legal advice.
Arm a saved definition before the next file, then see missing, unexpected, required, type, and allowed-value failures across every row.
Normal exports exclude source rows and raw values; only the separately disclosed issue ledger may include failing values.