sxsphinxstack

Skills / Starter / Validate a json schema

Starter Code skill

Validate a json schema

Validate a JSON Schema against intended instances, invalid boundaries, references, defaults, and consumer behavior.

Verify both what the schema accepts and what it rejects.

When to use

  • Use for configuration, APIs, events, manifests, stored documents, or generated data.
  • Do not assume default, format, or annotation keywords mutate or enforce data in every validator.

Procedure

  1. Record draft version, base URI, vocabulary, validator implementation, and consumer expectations.
  2. Resolve local and remote references without untrusted network fetching in production validation.
  3. Check object properties, required fields, additional properties, arrays, numbers, strings, nulls, enums, and composition.
  4. Create canonical valid examples for minimum and representative shapes.
  5. Create invalid cases for missing, extra, wrong-type, boundary, duplicate, and contradictory values.
  6. Test conditional logic, recursive references, formats, defaults, and unevaluated properties in the actual validator.
  7. Review compatibility impact on existing producers and consumers.
  8. Save schema, fixtures, validator version, and results together.

Related skills

  • Validate a json payload — Validate a JSON payload against syntax, schema, semantic invariants, size, encoding, and privacy requirements.
  • Validate a sql query plan — Validate a SQL query plan against representative parameters, statistics, data shape, resource use, and production-safe execution.
  • Validate a cron expression — Validate a cron expression against the exact scheduler dialect, time zone, calendar behavior, overlap, and operational intent.