How does this skill improve Claude's SQL accuracy?

By documenting company-specific 'gotchas,' such as specific test data filters, primary key relationships, and metric formulas, it prevents Claude from making generic assumptions that lead to incorrect results.

What is the difference between Bootstrap and Iteration mode?

Bootstrap mode creates a brand-new data context skill from scratch by exploring schemas, while Iteration mode updates an existing skill with new domains, metrics, or tables.

Does it require manual documentation of every table?

No, the skill uses automated discovery tools to pull schema details and then asks targeted questions to capture the 'why' and 'how' behind the most important tables.

Which databases does the Data Context Extractor support?

It is designed to work with major data warehouses including BigQuery, Snowflake, PostgreSQL, Redshift, and Databricks using standard MCP tools.

Data Context Extractor

Name: Data Context Extractor
Author: anthropics

byanthropics

•

8,000

•

データサイエンスとML

Generates tailored data analysis skills by capturing tribal knowledge, schema details, and business metrics from analysts.

The Data Context Extractor is a meta-skill designed to bridge the gap between raw data warehouses and Claude's ability to provide accurate business insights. By guiding users through a structured discovery process—including schema exploration, entity disambiguation, and metric definition—it generates a comprehensive, documented 'data skill' package. This ensures Claude understands your specific terminology, unique identifiers, and SQL dialects like BigQuery or Snowflake, effectively preventing common query pitfalls and aligning AI-generated analysis with established organizational practices.

主な機能

01Entity disambiguation to define complex business logic and relationships

02Standardized metric cataloging for accurate KPI calculations

03Automated schema discovery for BigQuery, Snowflake, Redshift, and Databricks

048,000 GitHub stars

05Automatic generation of structured documentation and reference files

06Interactive Bootstrap Mode to build custom data context from scratch

ユースケース

01Capturing undocumented tribal knowledge regarding data hygiene and exclusions

02Standardizing SQL query patterns and filtering logic across a data team

03Onboarding Claude to a complex corporate data warehouse with custom schemas

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add anthropics/knowledge-work-plugins data-context-extractor

For use in Claude.ai and ChatGPT

Download Skill