By name: data-normalizer metadata: version: "0.0.1" author: "ShuangShu" description: Normalize mixed-source records into canonical field shapes before validation, indexing, and downstream processing.
Data Normalizer
Overview
Use this skill to map heterogeneous payloads into a canonical schema and normalized value space.
Inputs
- Raw records from APIs/files/streams.
- Field mapping rules and default values.
Processing
- Normalize text and timestamps through dependency primitives.
- Coerce simple scalar types where safe.
- Track dropped and transformed fields.
Output Contract
- Return normalized records.
- Include field-level transformation summary.
- Flag lossy transformations explicitly.