dsh-json-types

You have a real JSON payload in hand and need a TypeScript interface, a JSON Schema, or a Pydantic model for it. Fields are scattered, nesting runs several levels deep, arrays contain objects, and an optional field is easy to miss. dsh-json-types exists for exactly this moment: paste the sample in and get a ready-to-use type definition out.

The plugin ships three tools: json_to_ts for TypeScript interfaces, json_to_schema for JSON Schema (draft-07), and json_to_python for Pydantic models. Inference covers primitive field types, nested objects with auto-generated named interfaces or models, array element types, and missing fields automatically marked as optional. Everything runs locally in Node with zero network calls or external services.

If you work with JSON data in front-end, back-end, or data-pipeline code and need quick type constraints for API responses, config files, or log samples, this plugin compresses repetitive typing into a single paste. Inference is sample-based, so edge cases such as a string that is really an enum or a date may need a light manual tweak after generation. Purely local, no external dependencies, MIT licensed, and suited to any developer who would rather not install a whole separate tool just for type definitions.