JSON Analyzer

What the analyzer measures

• File size in bytes, total nodes and maximum nesting depth.
• Counts of objects, arrays, strings, numbers, booleans and nulls.
• Total and unique property names.
• A per-field profile: how often each path appears, which types it holds, and sample values.

Data-quality checks

Structure statistics tell you the shape. The findings tell you where it is unreliable:

• Inconsistent types — the same field is a number in 99.9% of records and a string in the rest. This is the single most common cause of downstream crashes.
• Missing fields — a property that appears in only some sibling objects, which usually means it is optional.
• Null-heavy fields — a field that is null in a third or more of its occurrences.
• Empty strings and empty arrays, which often stand in for missing data.
• Duplicate identifiers — repeated values in a field named id, uuid, key, sku and similar.
• Mixed date formats — ISO 8601 in some records, MM/DD/YYYY or a Unix timestamp in others.
• Arrays that mix item types, which makes them awkward to map onto typed models.
• Very deep nesting, which makes a document hard to query and to render.

Findings are ranked, not alarmist

Every finding carries a severity. Warning means something is probably wrong and worth checking. Info means it is worth knowing but may be entirely intentional — plenty of well-designed APIs have optional fields and nullable columns. The analyzer describes what it observed; it does not claim your data is broken.

From findings to a fix

Inconsistent types and mixed date formats usually point at a producer that needs fixing. Optional fields and nullable columns should be reflected in your generated types and your JSON Schema — generate both from the same document so the contract matches reality.