verify-data-quality

To perform a data quality assessment for CRISP-DM, this Skill evaluates dataset completeness, correctness, consistency, and timeliness, then generates a Jupyter notebook with validation code and a markdown summary for the Data Understanding phase.

A data quality check for completeness and consistency systematically evaluates your datasets to identify missing values and logical rule violations, producing a machine-readable quality report and a notebook with validation results.

Yes, you can run data quality checks on multiple datasets simultaneously by providing their paths, optionally including reference documents to guide the validation process across completeness, correctness, consistency, and timeliness dimensions.

Validating dataset quality generates two artifacts: an executable Jupyter notebook at notebooks/2.4-data-quality.ipynb containing validation code, and a human-readable markdown summary at docs/crisp-dm/2-data-understanding/2.4-data-quality.md.

You need to run data quality validation during the Data Understanding phase of the CRISP-DM workflow to systematically assess dataset completeness, correctness, consistency, and timeliness before proceeding to data preparation.

No specific dependencies are required to generate a data quality report, as the Skill operates independently to produce the validation notebook and CRISP-DM summary document directly from your dataset inputs.