Analyze any CSV file through a conversational flow — no code, no Python, no external plugins required. Claude reads your file, profiles its structure, and guides you to the insights you need through a single clarifying question.
Designed for non-technical users who regularly work with exported data: sales reports, task lists, CRM exports, survey results, expense files.
- A CSV file accessible in your Cowork workspace (uploaded or in the selected folder)
- No additional setup, API keys, or tools needed
- Upload your CSV file to the Cowork workspace (or confirm it's in your selected folder)
- Trigger the skill with one of the phrases below
- Claude profiles your dataset and asks one question: what would you like to understand?
- Answer in plain language — Claude performs the analysis and delivers numbered insights
- Review the results and optionally explore follow-up directions
Example 1: Sales export
"Analyze my CSV" + uploads
sales-q1.csv
Claude reads the file, reports: 847 rows, 6 columns (date, region, product, units, revenue, rep). Asks: "What would you like to understand — summary statistics, trends over time, top/bottom items, or something else?" User replies: "Which product made the most revenue?" → Claude returns a table of products ranked by revenue with plain-language interpretation.
Example 2: Task list
"Help me understand this data" + points to
sprint-tasks.csv
Claude profiles the file: 63 rows, columns include status, assignee, priority, estimate. User says: "Show me how tasks break down by status." Claude returns a grouped count table and notes that 40% of tasks are still in "In Progress" with the sprint ending this week.
| Language | Trigger Phrase |
|---|---|
| EN | analyze my CSV |
| EN | analyze this CSV file |
| EN | help me understand this data |
| EN | what insights can you find in this spreadsheet |
| RU | проанализируй CSV |
| RU | разбери мой файл данных |
| RU | что в этом CSV |
| RU | помоги понять данные из таблицы |
Does:
- Read CSV files directly from your Cowork workspace
- Profile structure: row count, column types, sparse columns, missing data
- Perform natural-language analysis: summaries, trends, rankings, group comparisons
- Deliver business-readable insights without technical jargon
- Handle large files by sampling (>500 rows)
Doesn't:
- Generate charts or visual output (suggest pasting the table into Excel/Sheets for that)
- Execute Python or other code
- Connect to external databases or APIs
- Modify or overwrite your original file