AI Developer Tools

Claude Tool Schema Validator

Inspect one Claude client tool definition for its name, description, and input_schema object before using it in an API request.

✓ Every tool runs in your browser. We do not upload your input or output.

Claude Tool Schema Validator
Keyboard shortcuts

Ctrl/⌘ + Enter: process

Ctrl/⌘ + Shift + F: full screen

Esc: exit full screen

Ctrl/⌘ + F in editor: search

Schema check result
ReadyLines: 1 · Characters: 0 · Size: 0 KB · Time: — · Cursor: 1:1

How to use

  1. Paste one Claude client tool JSON object into the editor.
  2. Run the checker and read any structural error.
  3. Make sure input_schema is an object schema with a properties map.
  4. Compare with current Anthropic documentation and test your full tool list with the API.

Examples

Check a weather lookup

The example has a valid name, description, and empty object input schema.

Input
{"name":"get_weather","description":"Get weather","input_schema":{"type":"object","properties":{}}}
Output
Valid Claude client tool shape

Check a client tool’s basic contract

Claude client tools use a name, a description, and an input schema describing the arguments the model may provide. This page checks those fields and the basic JSON Schema object shape. It also checks the documented name character pattern and length. The goal is to catch common copy-and-paste mistakes before the tool is added to an application.

A structural check cannot evaluate whether a tool is safe to call, whether the description leads to good tool selection, or whether a particular API version accepts every nested schema keyword. Keep descriptions precise, define the arguments you actually handle, and test the complete request against the official Claude tool guide. This page does not call Claude or upload the definition. If you need to check strict JSON syntax separately, use the JSON Validator; if you are preparing an OpenAI function tool, use the related OpenAI checker.

A useful review is to compare each input property with the actual function implementation. Required fields should truly be required, names should be stable, and descriptions should say what the argument means. Good schema shape helps validation but cannot replace application-side checks.

Frequently asked questions

Does this send a request to Claude?

No. The JSON is checked locally in a browser worker.

Does a valid result guarantee acceptance?

No. API versions and model-specific limits may impose additional constraints.

Which tool type is checked?

This page checks a client tool with an input_schema, not every tool type Claude supports.

Are tool descriptions required?

Yes. A nonempty description helps explain the tool purpose and is required by this checker.

Related tools