Profiling Tables

This is the kind of profiling workflow you'd normally build yourself after the third time someone asks "what's actually in this table?" It runs through the standard checklist: column metadata, row counts, type-specific statistics (min/max/avg for numbers, length distributions for strings, date ranges for timestamps), cardinality analysis to spot IDs versus categories, and a data quality assessment covering completeness, uniqueness, and freshness. The output is a structured summary with a schema table, quality scores out of 10, and a list of potential issues. It's thorough without being clever, which is exactly what you want when onboarding someone to an unfamiliar dataset or debugging why a pipeline is producing garbage.

skills.sh installs: 1.1K
GitHub stars: 459
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add astronomer/agents --skill profiling-tables --agent claude-code

Installs into .claude/skills of the current project.

Files

Data Profile

Generate a comprehensive profile of a table that a new team member could use to understand the data.

Step 1: Basic Metadata

Query column metadata:

SELECT COLUMN_NAME, DATA_TYPE, COMMENT
FROM <database>.INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_SCHEMA = '<schema>' AND TABLE_NAME = '<table>'
ORDER BY ORDINAL_POSITION

If the table name isn't fully qualified, search INFORMATION_SCHEMA.TABLES to locate it first.

Step 2: Size and Shape

Run via run_sql:

SELECT
    COUNT(*) as total_rows,
    COUNT(*) / 1000000.0 as millions_of_rows
FROM <table>

Step 3: Column-Level Statistics

For each column, gather appropriate statistics based on data type:

Numeric Columns

SELECT
    MIN(column_name) as min_val,
    MAX(column_name) as max_val,
    AVG(column_name) as avg_val,
    STDDEV(column_name) as std_dev,
    PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY column_name) as median,
    SUM(CASE WHEN column_name IS NULL THEN 1 ELSE 0 END) as null_count,
    COUNT(DISTINCT column_name) as distinct_count
FROM <table>

SKILL.md on GitHub