What problem does it solve?
This Skill provides a fast, read-only way to inspect what a frontend dive will display by converting dive manifests and query results into human-friendly markdown pivot tables so stakeholders can review slices and spot data quality issues without opening a browser.
Core Features & Use Cases
- Manifest-driven SQL: Extract table, columns, where, filters, and groupBy from dive manifests and build precise SQL that mirrors the frontend's useDiveData contract.
- Read-only querying and rendering: Execute queries via the soria warehouse query tool, format time-series as columns and snapshots as entity rows, and render default and alternate slices as pivot tables.
- Quality checks and drift detection: Surface missing methodology, count verification checks from verifications metadata, and report manifest/data drift or empty slices with clear remediation advice.
- Use Case: Quickly show a product manager the default PDP slice and a segment-swapped slice for a given dive to validate presentation and surface any data anomalies before a deploy.
Quick Start
Run the preview skill to render the dive's default view and one alternate slice as a markdown pivot table using the manifest and soria warehouse query.