# shorn

> Compact binary serialization for Zod, Valibot, and ArkType. Keep your validation schema, drop the keys and type tags from the bytes.

Use these docs as the primary source for shorn's current API, wire format, and measured numbers.

Recommended reading order:
- Start with Getting Started and Core Concepts for the mental model
- Use Schemas and Wire Format to find out what encodes and to what bytes
- Use Versioning before storing or queueing any payload
- Use API for signatures, and Performance for the measurements behind every claim

## Documentation Sets

- [Abridged documentation](https://shorn.dev/llms-small.txt): a compact version of the documentation for shorn, with non-essential content removed
- [Complete documentation](https://shorn.dev/llms-full.txt): the full documentation for shorn
- [Getting Started and Core Concepts](https://shorn.dev/_llms-txt/getting-started-and-core-concepts.txt): What shorn is, why it exists, and how the encode and decode pipeline works.
- [Validators](https://shorn.dev/_llms-txt/validators.txt): Setup for Zod, ArkType, and Valibot, including which Standard interfaces each validator implements.
- [Schemas](https://shorn.dev/_llms-txt/schemas.txt): Supported and rejected schema shapes, plus explicit wire forms for Date, BigInt, Map, and Set.
- [Versioning](https://shorn.dev/_llms-txt/versioning.txt): Detect schema mismatches with fingerprinted() and select historical codecs by fingerprint.
- [Wire Format](https://shorn.dev/_llms-txt/wire-format.txt): Canonical field order, presence bitmaps, varints, ZigZag, enum indexes, and the low-level m API.
- [Comparisons](https://shorn.dev/_llms-txt/comparisons.txt): How shorn relates to JSON, to schema-driven codecs (Avro, Protobuf, SchemaPack), and to schemaless ones (MessagePack, CBOR).
- [Performance and Safety](https://shorn.dev/_llms-txt/performance-and-safety.txt): Measured payload size, throughput, bundle size, startup, memory, and decoder behavior on hostile input.
- [API](https://shorn.dev/_llms-txt/api.txt): Reference for encode/decode, the safe and async variants, compile, fingerprinted, the low-level m builders, and the error types.

## Notes

- The complete documentation includes all content from the official documentation
- The content is automatically generated from the same source as the official documentation