validation-quarantine

Data validation with Zod schema and custom quality scoring categorizes records into valid, quarantined, and invalid sets. Suspicious records move to a quarantine store for manual review, preventing pipeline blockage while ensuring data conformity.

Data quarantine manages a store for records scoring below a quality threshold. Features include approval, rejection, and auto-rejection of stale items, allowing manual review of suspicious data without halting the ingestion pipeline.

Quality scoring assigns a score based on custom domain rules during batch processing. This identifies potentially problematic data, moving records with scores below a threshold to quarantine for later review.

Validate input against expected formats using Zod schemas, score completeness and relevance with custom rules, and quarantine entries that seem incomplete or unusual for a human to review later without halting the pipeline.

Yes, the skill uses TypeScript and Zod schemas for schema validation. It ensures incoming data conforms to a defined structure and applies custom quality scoring to categorize records efficiently.

The quarantine mechanism manages a store for suspicious records with features for approval, rejection, and auto-rejection of stale items. This allows human review of edge cases without blocking the data ingestion pipeline.