bmad-distillator

Generate lossless distillates from mixed source documents for LLM workflows.

Updated Dec 23, 2025

Search Tags:#documentation #validation #llm #compression #semantics #token-budget #distillate

Ready to run

npx skills add https://github.com/joekhosbayar/go-mighty --skill bmad-distillator-joekhosbayar

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bmad-distillator
Source: https://github.com/joekhosbayar/go-mighty/tree/main/.gemini/skills/bmad-distillator
Command: npx skills add https://github.com/joekhosbayar/go-mighty --skill bmad-distillator-joekhosbayar

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Distillator provides lossless compression of source documents into a single distillate or semantically split set for efficient downstream LLM consumption, preserving all facts, decisions, constraints, and relationships.

Core Features & Use Cases

Lossless document distillation: compresses multiple sources into a dense, token-efficient format.
Supports single or semantically split distillates, token budgeting, and round-trip validation.
Stage-driven workflow: analyze sources, compress, verify, and optionally reconstruct for validation.
Facilitates downstream workflows like PRD creation, architecture design, or knowledge synthesis.

Quick Start

Use the distillator to convert source documents into a compact distillate by providing source_documents and optional downstream_consumer, token_budget, and output_path.

Frequently Asked Questions about bmad-distillator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I compress multiple markdown and yaml documents for LLM token budgeting?▼

Lossless document distillation compresses multiple source documents into a dense, token-efficient format. It analyzes mixed sources like markdown, txt, yaml, and json, then outputs a single distillate or semantically split set while preserving all facts and relationships.

Can I set a specific token budget for downstream LLM workflows?▼

Yes, you can specify an optional token_budget to constrain the output size for downstream LLM workflows. This ensures the generated distillate fits within your model's context window while retaining all critical facts, decisions, and constraints.

What is lossless document distillation and when do I need it?▼

Lossless document distillation compresses source documents into a dense format without losing facts, decisions, or relationships. You need it when preparing large mixed-format knowledge bases for efficient downstream LLM consumption.

How do I validate that a distillate retains all original document semantics?▼

You validate distillate semantics by using the round-trip validation stage. Triggered by a flag, this stage verifies the compressed output by optionally reconstructing it to ensure all original facts and relationships are preserved.

Does document distillation work with mixed json and text source files?▼

Yes, distillation works with mixed source files including json, yaml, txt, and markdown. It analyzes these varied formats and compresses them into a unified distillate or semantically split outputs for downstream LLM workflows.