Explore, clean, and summarize CSV datasets with natural language queries.
CSV Analyzer lets you interrogate CSV files using plain English. Ask 'what are the top 10 customers by revenue' or 'show me rows where status is overdue and amount > 1000' and the skill translates your query into the appropriate column filters, aggregations, and sorts.
The skill handles common data quality issues automatically: it detects and parses date formats, normalizes currency strings, trims whitespace, and identifies likely numeric columns stored as text. A `/csv summary` command provides a quick profile — row count, column types, null percentages, unique values, and basic statistics (min, max, mean, median) for numeric columns.
Output is formatted as Markdown tables for small results or written to a new CSV for large outputs. The skill reads and writes local files only — no data leaves your machine. Supports files up to ~500MB with streaming row processing.