bmad-distillator

Lossless document compression strips human-readable overhead to create dense, token-efficient files, preserving all facts and relationships for LLM consumption. This solves context window limits by optimizing source documents into a hyper-compressed format.

Lossless distillation extracts essential information from source documents into a dense format optimized for LLM processing. It strips human-readable overhead while preserving all facts, decisions, and relationships, ensuring complete round-trip reconstruction.

Yes, lossless distillation handles multiple input files, combining them into a single dense document. You can distill a large product brief and discovery notes together for an LLM to use as sole context for generating a PRD.

Document distillation supports optional downstream consumer context and token budgets. By specifying these parameters, the compression process formats the dense output to fit specific LLM constraints while maintaining lossless information reconstruction.

Yes, lossless document compression includes validation for lossless round-trip reconstruction. This ensures all facts, decisions, and relationships from source documents are fully preserved during the token-efficient information extraction process.