toon-format

To reduce token usage, convert JSON data into Token-Oriented Object Notation (TOON) format. TOON provides a compact, LLM-friendly data serialization method that maintains data integrity while significantly lowering token counts in prompts.

The best way to serialize structured data for large language models is using TOON format. It optimizes data representation for LLM token efficiency, allowing you to process more information within context limits compared to standard JSON.

TOON format syntax provides specific, clear guidelines for structuring arrays, objects, and quoting conventions. These rules ensure compact data representation while preserving the structural integrity needed for accurate LLM parsing.

Yes, you can convert existing JSON files to TOON format automatically. The conversion process adheres to specific TOON syntax rules, transforming complex JSON into a compact structure that minimizes LLM token consumption.

You need to use a TOON file for knowledge sharing when capturing and distributing organizational learnings. Creating an `org-patterns.toon` file defines a specific, token-efficient format for sharing this structured knowledge with LLMs.

TOON format is optimized specifically for LLM token efficiency and requires adherence to its syntax rules for arrays, objects, and quoting. It is specialized for compact data representation rather than general-purpose data serialization.