What structural consistency means
A record-by-record view of JSON array shape
Structural consistency describes how closely objects in an array share the same fields and value types. This analyzer measures each record against the dominant observed structure at its normalized array path.
API response QA
Check saved responses before integration work
Compare repeated API records for missing properties, unexpected fields, and changing types. The object-level report makes a variation traceable to the exact record and normalized path.
ETL validation
Inspect records before an import or transformation
A field that changes from a number to a string can break a downstream load even when the JSON itself is valid. Review measured field presence and type consistency before moving data.
Data quality
Separate syntax validity from structural uniformity
Valid JSON can still contain irregular records. The consistency score summarizes observed shape and type agreement without inferring intent, changing data, or claiming schema compliance.
Examples
Understand each reported difference
| Finding | Example |
|---|---|
| Missing field | active appears in the dominant structure but not object 3. |
| Extra field | segment appears only outside the dominant structure. |
| Type mismatch | id is usually a number but is a string in one object. |
Inspection utilities
Continue with related UtilityDock inspection workflows
FAQ
Questions about JSON structural consistency
Does UtilityDock upload or store my JSON?
No. Your files never leave your computer. All processing happens locally inside your browser.
How is the consistency score calculated?
Each record is compared with its array's most common observed field structure and the most common observed type for each field. The score is the percentage of records with no measured variation.
What is a missing or extra field?
A missing field belongs to the dominant observed structure but is absent from a record. An extra field appears in a record but not in that dominant structure.
How are nested arrays handled?
Object arrays found at the same normalized path are analyzed together. Numeric indexes are represented as [*] so repeated nested arrays can be measured consistently.
Does this tool repair data or generate a schema?
No. Version 1 is read-only. It reports observed differences without changing JSON, repairing records, generating schemas, or calling APIs.