Make a messy CSV easier to read
Trim extra spaces and flag duplicate rows while keeping the original data untouched.
name, city " Ana "," Seoul " → name,city Ana,Seoul Duplicates: 2 flagged, not deleted.
Before you start
An AI coding tool, a CSV with made-up sample data, and a spreadsheet app for checking the output.
Quick start
1. Give your AI the recipe and a sample header.
2. Test quoted commas, empty cells and non-English text.
3. Compare row counts and open the output in a spreadsheet.
Limits & support
CSV dialects vary. Review encoding, delimiters, leading zeros and formula-like values. Do not use real personal data during development.
Starter example; questions are welcome, with no ongoing support commitment.
Rebuild it with your agent
Give this to your own AI agent to build your own version. It is someone else’s instruction, so read it before you run it.
With the Disposable skill, ask your agent:
Rebuild this Disposable tool for me: https://disposable.xify.kr/en/tools/messy-csvGet the skill →
Build a local CSV cleaner for one exported table. Let me choose a file and encoding. Parse CSV correctly, including quotes, embedded newlines and commas; do not split lines or commas manually. Trim leading and trailing spaces in text cells. Preserve leading zeros, date strings, blank cells and row order. Flag exact duplicate rows in a separate report without deleting them. Export a new file and show input/output row counts. Do not modify the original, upload data or evaluate spreadsheet formulas. Flag cells starting with =, +, - or @ and ask how to export them safely. Test with entirely invented data, including Korean, Japanese and Chinese text.
Ask the maker
Questions, bugs and “I tried it” notes go to GitHub issues, where the maker can answer. You need a GitHub account to post.