danielrosehill/Claude-Synthetic-Data-Plugin

evaluate-quality

Run fidelity, utility, and privacy diagnostics on a synthetic dataset against the real source, generating a markdown report.

Idioma del texto original: inglés

real-to-synth-llm

Transform real text records into synthetic counterparts that preserve semantics while changing specifics, using Claude.

Idioma del texto original: inglés

replace-pii

Anonymise a real dataset by replacing PII columns with realistic Faker values, preserving structure and referential integrity.

Idioma del texto original: inglés

setup-workspace

Initialize a synthetic data workspace directory with standard folders, Python requirements, and venv setup.

Idioma del texto original: inglés

tabular-from-real

Fit a generative model on real tabular data and sample synthetic rows preserving correlations and distributions.

Idioma del texto original: inglés

tabular-from-schema

Generate a synthetic tabular dataset from a JSON schema describing columns, types, and Faker providers.

Idioma del texto original: inglés

text-records-llm

Generate synthetic text records (support tickets, reviews, medical notes, etc.) using Claude based on a persona pool and schema.

Idioma del texto original: inglés

tools-reference

Reference card of open-source synthetic data tools, when to use each, install commands, and design patterns.

Idioma del texto original: inglés