Cookbook
Pick a role and a goal, then jump to a verified command path.
Privacy / compliance
| Goal | Start here |
|---|---|
| Find PII in a CSV | PII detection |
| Scan a database for PII | Database scanning |
| Restrict rules to English PII | metacrafter scan file users.csv --contexts pii --langs en |
| Push labels into DataHub | Catalog export |
Data engineers
| Goal | Start here |
|---|---|
| Classify columns in Parquet / Excel | File scanning |
| Scan every file in a directory | scan bulk |
| Use Metacrafter from Python | Python SDK |
| Run a shared classifier API | API server |
Analysts / stewards
| Goal | Start here |
|---|---|
| Understand what a column is | Semantic labeling |
| Look up a datatype | Registry |
| See loaded rules | rules |
| Statistics without classification | metacrafter scan file data.csv --stats-only |
Rule authors
| Goal | Start here |
|---|---|
| Add a YAML rule | Custom rules |
| Install the extended pack | pip install 'metacrafter[rules]' |
| Filter by country | --country-codes us,ca |
| Hybrid rules + LLM | LLM classification |
First commands
# Human-readable table
metacrafter scan file data.csv --format short
# PII-focused JSON report
metacrafter scan file users.csv --contexts pii --format full --output-format json -o report.json
# Database
metacrafter scan sql "sqlite:///app.db" --format full -o db.json
# Inspect rules
metacrafter rules list --output-format json