verifying-training-data-erasure

What problem does it solve?

This Skill helps teams verify and document that a subject’s data has been erased across the full ML pipeline after a right-to-erasure or right-to-amend request.

Core Features & Use Cases

• Multi-stage erasure verification: Checks raw datasets, embeddings and vector stores, model weights, inference caches, and backups or replicas for lingering subject data.
• Audit-ready proof bundles: Produces structured evidence with commands run, outputs, sign-offs, residual-risk statements, and retention dates for privacy review.
• Compliance workflows: Supports GDPR, CCPA, HIPAA, and similar DSR cases where trained models or retrieval systems may still retain subject-derived information.

Quick Start

Ask this Skill to walk a specific subject through erasure verification and generate a DSR proof bundle for the relevant regulation.