Browse skills

Analyze SQL queries and EXPLAIN plans to identify performance bottlenecks, then produce optimized rewrites with index recommendations. Supports PostgreSQL, MySQL, SQLite, and SQL Server dialects. Covers full table scans, N+1 patterns, missing indexes, subquery flattening, join optimization, and query plan interpretation. Returns optimized query, index DDL, and estimated improvement.

A skill for transforming data between formats with configurable options.

Extract structured data from any unstructured text — invoices, emails, contracts, resumes, support tickets, medical records, legal filings, receipts, meeting notes. Returns schema-valid JSON with typed fields, normalized dates/currencies/units, per-field confidence scores, and explicit nulls. Handles multi-entity extraction, nested objects, and array detection. Zero hallucination guarantee with source spans.

Inspect a SQL statement before it runs. Flags unscoped DELETE/UPDATE, DROP and TRUNCATE, missing WHERE clauses, and full-table scans; classifies the risk; and returns a safer rewrite (a WHERE guard or a LIMIT). Stops an agent from nuking a table by accident.

Run a persistent Jupyter kernel for interactive data science: execute cells, keep state across steps, inspect dataframes, and render plots — without re-running the world.