Pasted CSV analysis that checks the data before answering
Answers a business question about a CSV you paste or upload. It profiles the data first, states how each number was computed, separates findings from guesses and lists caveats, so the answer can be trusted.
Prompt in Data by AISkillHub Library
What it does
Asking a chat model "what does this data say?" often produces confident numbers that were never actually computed. This prompt makes the analysis disciplined: it profiles the data first (rows, columns, types, missing values, date range), states every calculation it performs, uses a code execution tool when one is available, and keeps findings separate from interpretation.
**Use it when** you have an export (sales, signups, tickets, survey results, inventory) and a specific question to answer for someone else, such as "which region drove the drop?" or "is response time getting worse?"
**What you get** - A data check: shape, column types, date range, missing or odd values and how they were handled. - A direct answer to your question in two to four sentences, with the key numbers. - A supporting table and, if useful, a description of the one chart that best shows the result. - The method: each calculation in plain words, including filters and definitions. - Caveats and alternative explanations. - Two or three next questions the data could answer.
**Works best with a code tool.** ChatGPT, Claude and Gemini each offer a way to run code on uploaded files, depending on your plan and settings. When available, the prompt tells the model to compute with code rather than estimate. Without one, keep pasted data small (a few hundred rows) and verify key figures in your spreadsheet.
**What it will not do** - It will not report a number it did not compute from the data. - It will not claim causation from a correlation. - It will not silently drop rows. Exclusions are counted and explained. - Remove personal or confidential data your organisation does not allow in chat tools before pasting.
How to use it
## How to use
1. Open `prompts/csv-data-analysis.md` and copy everything below the title line. 2. Replace each `{{variable}}` with your own content. Paste long material (notes, transcripts, tables) directly in place of the variable; keep the surrounding tags so the model can tell inputs apart. 3. Start a new chat in ChatGPT, Claude or Gemini and paste the filled prompt as your first message. 4. Read the output against the quality checks at the end of the prompt before you use or send it.
## Variables
- `{{question}}`: the specific business question to answer. - `{{csv_data}}`: the CSV text with its header row, or write `attached` if you upload the file. - `{{data_notes}}`: what a row represents, units and currency, known quirks, and what is excluded. - `{{audience}}`: who will read the answer, e.g. `regional sales managers`.
If you do not have a value for an optional variable, write `none` rather than deleting it, so the prompt still reads correctly.
## Example
`prompts/csv-data-analysis.example.md` shows the prompt with every variable filled and a trimmed sample of the output, so you can see the expected shape before you run it on real material.
## Reusing it
To reuse it often, save the prompt (with the variables left in) as the instructions of a Claude Project, a custom GPT or a Gemini Gem, then paste only the inputs in each new chat.
Files
Save each file at the path shown, relative to your project root.