Analyze Real Data Files Without Building a Separate Workflow
Load structured files, inspect the shape, surface patterns, and answer the question you actually care about.
A lot of lightweight analysis work dies in the gap between "I have a file" and "I need a clean answer." The Data Analysis skill closes that gap by combining file access, lightweight command-line analysis, and a reusable reporting workflow.
Best For
- CSV exports from analytics, revenue, surveys, or support systems
- JSON files, logs, or ad hoc operational datasets
- Fast directional analysis before you open a notebook or BI tool
Why This Workflow Wins
- It starts with the real file instead of forcing you to paste rows into chat.
- Bash support helps with quick calculations, grouping, and sanity checks.
- It is useful for questions that are too big for manual scanning but too small for a full analysis project.
Use Cases
Find the pattern in a CSV export
Drop in a report, ask a concrete question, and get the shape of the data plus the answer that matters.
Analyze this CSV and tell me which segments are growing fastest month over month.
Sanity-check an operational dataset
Use the skill to identify outliers, missing values, suspicious rows, or changes you did not expect.
Summarize this file, flag outliers, and tell me what looks wrong or incomplete.
Turn raw exports into a stakeholder summary
Ask for a short executive readout after the initial analysis so the output is ready to share.
Analyze this export, then give me a short executive summary with the 3 biggest findings.
System Prompt
You are a data analyst. When given a data file, understand its structure, then provide: a summary, key statistics, notable patterns or outliers, and answers to any specific questions.
Example Output Shape
A CSV file with 1,000 rows of sales data
Dataset summary, top products by revenue, monthly trends, average order value, and anomalies flagged.
Tips
- Attach the data file directly rather than pasting it.
- Ask specific questions for better output.
- For large files, mention the key columns you care about.