importing-data

What problem does it solve?

This Skill provides a structured, auditable pipeline to import CSV data into a SQLite database, eliminating ad-hoc imports and ensuring consistent, reproducible results from discovery through quality reporting.

Core Features & Use Cases

• Phase 1: CSV Discovery & Profiling: detect encoding, delimiter, headers, and capture representative samples; generate a detailed profile to inform schema design.
• Phase 2: Schema Design & Type Inference: infer column data types, map NULL representations, and propose a raw_[table_name] CREATE TABLE with justifications.
• Phase 3: Basic Standardization: define rules for date normalization, numeric cleaning, whitespace handling, and NULL mappings to ensure clean raw data.
• Phase 4: Import Execution: create the table, import the CSV with applied standardization, verify row counts and sample data, and run validation checks.
• Phase 5: Quality Assessment & Reporting: delegate quality checks to sub-agents to identify NULL patterns, duplicates, outliers, and potential foreign-key relationships, compiling a comprehensive quality report.

Quick Start

1. Start a new analysis session and run Phase 1 to generate 01-csv-profile.md.
2. Review Phase 1 output, then proceed to Phase 2 to draft the schema, followed by Phase 3 standardization rules.
3. Execute Phase 4 import and complete Phase 5 quality reporting using the provided templates.