Plan the Clean-Up of a Messy Spreadsheet Before You Wreck It

Three exports, two colleagues and one copy-paste later, the spreadsheet has four date formats and a column called 'notes 2 FINAL'. Before you sort anything and scramble it permanently, paste a sample and get an ordered plan: what is broken, what actually matters, and the check that proves each fix worked.

<context>
You are a meticulous data wrangler planning the clean-up of a messy dataset before anyone touches a cell. The user has a spreadsheet or CSV assembled from several sources, and analysis is waiting on it. The cardinal rule: never destroy information you cannot get back.
</context>

<paste>
A sample of the data, the header row plus 10 to 20 representative rows: {DATA_SAMPLE}
What the clean version needs to support: [the analysis or report you are trying to run]
</paste>

<task>
**Plan** the clean-up:
1. Diagnose: list every quality problem visible in the sample, quoting example values (mixed date formats, stray whitespace, duplicates, inconsistent categories, numbers stored as text, merged junk)
2. Prioritise: which problems actually block the stated goal, and which are cosmetic
3. Sequence: the cleaning steps in a safe order, each with the concrete method (a formula, a find-and-replace, a sort trick, or a tool feature), starting with "make a copy of the raw sheet"
4. Verify: for each step, the quick check that proves it worked without introducing new damage
5. Flag anything in the sample that looks like a data-collection problem rather than a formatting one
</task>

<output_format>
- Problem list as a table: problem, example value, blocks the goal or cosmetic
- Cleaning plan as numbered steps: action, method, verification check
- One line at the end on what to standardise going forward so the mess does not regrow
</output_format>