bmad-distillator

Compress source documents into lossless distillates with frontmatter metadata.

Updated Mar 26, 2026

Search Tags:#llm #compression #workflows #document #token-efficiency #distillation

Ready to run

npx skills add https://github.com/jaydubya818/New_baseline --skill bmad-distillator-jaydubya818

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bmad-distillator
Source: https://github.com/jaydubya818/New_baseline/tree/main/skills/bmad/core/bmad-distillator
Command: npx skills add https://github.com/jaydubya818/New_baseline --skill bmad-distillator-jaydubya818

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Lossless distillation compresses multiple source documents into a single or semantically split distillate that preserves every fact, decision, constraint, and relationship while removing redundancies for downstream LLM workflows.

Core Features & Use Cases

Analyze input sources to determine grouping and splitting strategy.
Compress information into a distillate with optional semantic splitting and frontmatter metadata.
Verify format and (optionally) perform round-trip validation to reconstruct originals from the distillate.
Designed for token-efficient, deterministic AI pipelines that require context-rich inputs.

Quick Start

Provide input paths to the distillator and include an optional --validate flag to verify lossless reconstruction.

Frequently Asked Questions about bmad-distillator

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

FAQPage Schema

How do I compress source documents for token-efficient LLM workflows?▼

Document distillation compresses multiple source documents into a single or semantically split distillate that preserves every fact and relationship while removing redundancies for token-efficient downstream AI workflows. You provide input paths and optionally use a --validate flag.

What is lossless document distillation for LLM inputs?▼

Lossless document distillation is the process of compressing information into a distillate that preserves every fact, decision, constraint, and relationship. It optionally performs round-trip validation to reconstruct originals from the distillate to verify no data was lost.

Can I split compressed documents into semantic sections for targeted analysis?▼

Yes, semantic splitting is a core feature of document distillation. You can compress information into a distillate with optional semantic splitting, allowing you to generate sections specifically for targeted analysis, design, and decision pipelines.

What's the best way to verify that document compression preserved all original facts?▼

The best way to verify lossless compression is using the optional round-trip validation feature. It reconstructs originals from the distillate to verify format and confirm that every fact, decision, and relationship was preserved during the document distillation process.

Do I need to add metadata to track document provenance in AI pipelines?▼

You do not need to add metadata manually because the distillation process automatically includes frontmatter metadata for provenance. This ensures deterministic AI pipelines can trace compressed distillates back to their original source documents.

Why does my LLM workflow lose context when processing multiple large documents?▼

Large documents exceed token limits and cause context loss in LLM workflows. Document distillation solves this by analyzing input sources, determining grouping and splitting strategy, and compressing information into token-efficient, context-rich distillates with frontmatter metadata.