Toon Parse Mcp

This server tackles token bloat by converting verbose data formats into TOON (Token-Oriented Object Notation) and stripping comments from code files. It exposes two tools: optimize_input_context for converting JSON, XML, YAML, and CSV into compact TOON format, and read_and_optimize_file for removing inline comments and whitespace from local code while preserving docstrings and structure. It also provides a mandatory-efficiency resource that prompts the LLM to prioritize these optimization tools. Reach for this when you're bumping against context limits with large data payloads or verbose codebases, especially in token-constrained environments where every token counts toward API costs or context windows.