CSV/Dataset Explainer

I'll paste a CSV header + a few sample rows. Tell me what this dataset actually is.

Output:

**What this dataset appears to be** — 1-2 sentences. What kind of records, what domain, what time period if visible.

**Column-by-column read** — for each column:
- What it probably represents
- The data type
- Any weirdness (mixed types, likely encoding issue, suspiciously round numbers)
- Confidence: high / medium / low

**Questions you could answer with this data** — 5-8 specific analytical questions. NOT "what patterns exist" — real questions like "which region has the highest average order value" or "does churn correlate with plan tier."

**Questions you CAN'T answer with this alone** — things that would require joining to another dataset (which one), a time dimension you don't have, etc.

**Suggested first three analyses** — the highest-value places to start.

Rules:
- Never claim certainty about column meanings if the header is ambiguous. Mark as low confidence.
- If the sample suggests PII, flag it and remind me to handle carefully.
- If the data appears to be a subset (looks like a query result), note it.

## How to use it

Paste this prompt, then paste your CSV header + 5-10 sample rows. Use the output to write your first analysis plan before you touch a chart or formula.

## Why it works

Analysts often waste hours re-deriving what a column means from scratch. A structured read-out gets you oriented fast, and the "can't answer" section stops you from chasing a question the data won't support.