What problem does it solve?
Deep-dive data analyses often produce one-off answers that cannot be reproduced, audited, or shared. This Skill structures BigQuery investigations into a persistent directory of artifacts so every insight is backed by a saved query, visualization, or table.
Core Features & Use Cases
- Structured analysis scaffolding: Creates an
analyses/ directory with assets/queries and assets/visualizations subdirectories plus a README as the main readable document.
- Reproducible artifacts: Saves every material BigQuery query as a commented
.sql file, generates charts via Python scripts with output images, and stores summary tables as .csv files.
- Rigorous methodology: Enforces an explicit plan, a Problem Statement, and carefully controlled cohort definitions before querying, then produces a final README with TL;DR and Key Takeaways.
- Use Case: When asked why paid users retain better than free users, the agent scaffolds an analysis folder, defines cohorts controlling for tenure and observation windows, saves each retention query and curve, and delivers a shareable README report.
Quick Start
Ask the agent to do a deep-dive analysis of a product question using BigQuery and produce a reproducible analysis directory with saved queries and visualizations.
