awesome-serialization

Agent2Agent Protocol. Google's open protocol for agent-to-agent communication and interoperability. JSON based. Textual.

Boundary AI Markup Language. Domain-specific language for defining LLM function signatures with type-safe structured output. Textual.

Lightweight markup widely used as the native "language" of LLM input/output. Highly token-efficient vs HTML/XML. Textual.

Model Context Protocol. Anthropic's open standard for connecting LLM agents to tools and data sources. JSON-RPC based. Textual.

Token-Oriented Object Notation. Compact, schema-aware JSON alternative achieving 30–60% token savings for LLM prompts. Textual.

Indentation-based format often more token-efficient than JSON for LLM contexts due to lack of braces/quotes. Textual.

Python-based Directed Acyclic Graphs for workflows.

Specification for describing analysis workflows and tools in a way that makes them portable and scalable across a variety of software and hardware environments.

Scientific workflow management, compatible with WDL and CWL.

Scalable and reproducible scientific workflows.

Coursera course on graph data manipulation.

Workflow Description Language for genomics and scientific workflows.

Cross-language columnar data format optimized for analytics workloads. Binary.

Scheme embedded, dynamic rich data structures. Textual/Binary.

Transactional storage layer for big data workflows. Binary.

Protocol Buffers suitable for larger datasets. Binary.

Open table format for large datasets. Binary.

Row storage with skip scan parsing. Structured, schema embedded. Amazon. Textual/Binary.

Advanced Scientific Data Format for astronomy and beyond. Binary/Textual.

n-dimensional datasets, complex objects, with schema. Efficient I/O. Binary.

Self-describing, machine-independent data format for scientific data. Binary.

Numpy arrays, cell sparse metadata. Binary.

Scalable storage of n-dimensional arrays. Binary.

OpenAPI equivalent for event-driven and message-driven architectures. Textual.

Binary schemeless JSON encoding. Binary.

High-performance, schema-based data interchange format. Binary.

Concise Binary Object Representation. Schema-free. Binary.

CNCF specification for describing event data in a common way. Textual/Binary.

Modern RPC framework compatible with gRPC, with HTTP/1.1, JSON, and browser support. Binary/Textual.

JavaScript implementation of Apache Avro. Textual.

High-performance binary serialization for Rust.

BSON serializer. Binary.

RAM to Disk serialization. Dart-specific. Binary.

Go's built-in serialization format for arbitrary data structures. Binary.

RAM to Disk serialization. Binary.

General theory of functions. Axiomatic foundation for mathematics, as an alternative to set theory.

Study of efficient serialization techniques for scalability.

Research on optimizing graph serialization.

Studies types, which informally are attributes that objects can possess.

Apple's on-device ML model format. Binary.

Quantized model format for llama.cpp/ggml. The de facto standard for local LLM inference. Binary.

TensorFlow graphs. Binary.

Intermediate representation for machine learning computations. Textual/Binary.

Apple's ML framework format, safetensors-based. Optimized for Apple Silicon. Binary.

Open Neural Network Exchange. Interoperability focused. Binary.