Wrangle_ your data_
A data workbench that runs entirely offline. Drop in a messy CSV, watch the preview update live, and export a clean file.
about Wranglify
Wranglify is the desktop version of Unfckr.app, a data workbench that runs entirely in your browser. Drop in a messy CSV or a wall of nested JSON, compose a pipeline of transformations, watch the preview update live, and export a clean file. No servers, no accounts, no uploads - verifiably. It works fully offline after your first visit - airplane mode is a feature, not a failure.
how to use it
ingest
Drop a .csv / .tsv / .json / .ndjson / .xlsx anywhere, paste straight from your clipboard, or hit browse. An Excel workbook opens by sheet, reading values as the file stores them - so leading zeros and long ID numbers arrive intact, and dates are dates rather than 45678. Big files stream: a 500 MB CSV previews in milliseconds.
inspect
The left panel counts the whole file, profiles every column, and raises a watchlist of what will fight you. Empty or constant columns, mostly-null columns, case-colliding values, duplicate headers. Quarantined rows are shown with reasons, never silently dropped.
unfck
Drop, rename, filter, find & replace, trim/case, dedupe, sort, scrub PII (mask / salted hash / drop, with Luhn + SSN validity checks), fix mojibake (byte-verified), flatten nested JSON, and paint a regex from examples. A step that breaks is flagged, never a crash.
export
CSV, JSON, NDJSON, or a real .xlsx workbook - your whole file, not the preview. CSV writes the way you need it - delimiter, LF or CRLF, and the BOM. Or skip the workaround entirely and export xlsx: a real workbook has no encoding to guess and no delimiter to pick.
recipes
Your steps stay with the file they were built for. Open a different file with the exact same columns and you are offered them rather than having them applied behind your back. Save a pipeline as a file and re-apply it to next week's dump, or share it as a link (in the URL fragment, safely offline).
engine
Need joins, a real full-file sort, or group-bys? Your pipeline's output loads into an actual SQLite database inside your browser. The ⌗ relationship map finds the keys between tables and measures them before you join so you know if rows will drop or multiply.
every tool in the box
Web version is free, private, works offline. Desktop is paid and also privarte, but includes terminal command scripts and batching for management thousands of docs at once.