JSON Schema validator

Why the error path matters more than the error

Knowing a payload is invalid is not useful on its own. What you need is the location: which field, at which depth, in which array element. Every issue reported above carries its JSON pointer, so you can go straight to the line that broke rather than reading the whole document.

The failure a status code will never show you

Consider an endpoint returning a user object. One day a migration makes id a string instead of an integer. The API still answers, still returns 200 OK, still looks perfectly healthy on any uptime dashboard. Your mobile client crashes on parse, and you find out from a user.

A schema catches that in one check. It is the difference between monitoring that the server is alive and monitoring that it is correct.

Writing a schema that is worth having

• List the fields your code genuinely depends on in required, not every field that happens to be there.
• Type the values. integer instead of number when it is an id, enum when the set is closed.
• Use minItems on collections that should never be empty. Empty arrays are the most common silent failure.
• Leave volatile fields such as timestamps and request ids unconstrained, or you will be paged by your own schema.

From a manual check to a continuous one

Once the schema is right, it belongs in a check that runs every few minutes rather than in a browser tab. Safeship stores it against the endpoint and verifies every response. If you use Claude Code or Cursor, the Safeship MCP server can read your endpoint, propose the schema and create the check without you writing it by hand.

Common questions

How do you validate JSON against a schema?

Write a JSON Schema describing the shape a valid document must have, then run both through a validator. The schema states which fields are required, what type each one holds and any constraints on its value. The validator returns the list of places where the document and the schema disagree, each with the path to the value that failed.

What is JSON Schema used for?

Declaring the contract a JSON document has to satisfy, so it can be checked automatically rather than by reading it. Common uses are validating API responses and request bodies, checking configuration files before an application starts, and generating documentation or types from a single definition.

Is my data sent anywhere?

No. Validation runs entirely in your browser. Nothing in either box leaves your machine, which makes it safe to paste a real production payload.

Which JSON Schema drafts are supported?

Draft-07 and the 2019-09 and 2020-12 keywords supported by Ajv, including format checks such as email, uri, date-time and uuid.

Why validate a schema instead of just checking for a 200?

Because the expensive failures return 200. A missing field, a number that became a string or an array that came back empty are all successful HTTP responses carrying broken data.

Can I run this validation against my live API automatically?

Yes, that is exactly what Safeship does. You attach a schema to an endpoint and every check verifies the response against it, then alerts you when it stops matching.