bmad-distillator

Extract facts, decisions, constraints, and relationships into lossless distillates with frontmatter metadata.

Updated Apr 25, 2026

Search Tags:#workflow #llm #compression #document #token #information-extraction #distillate

Ready to run

npx skills add https://github.com/ldk-hub/broke-shopping --skill bmad-distillator-ldk-hub

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bmad-distillator
Source: https://github.com/ldk-hub/broke-shopping/tree/main/.claude/skills/bmad-distillator
Command: npx skills add https://github.com/ldk-hub/broke-shopping --skill bmad-distillator-ldk-hub

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Distillator solves the challenge of converting multiple source documents into a single, lossless distillate that preserves every fact, decision, constraint, and relationship while minimizing token overhead for downstream LLM workflows.

Core Features & Use Cases

Hyper-compact, lossless distillates that maintain original information fidelity for subsequent AI tasks
Supports single-file or semantically split distillates to fit downstream context windows
Frontmatter metadata and provenance annotations enable round-trip validation and traceability

Quick Start

Provide source_documents and optional downstream_consumer to distill; the tool will output a distillate saved next to the primary source by default.

Frequently Asked Questions about bmad-distillator

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

FAQPage Schema

How do I compress documents for LLM context without losing information?▼

You can compress documents losslessly by extracting every discrete fact, decision, constraint, and relationship into a single dense distillate or semantically split sections, minimizing token overhead while preserving original information fidelity.

What is lossless distillation for compact LLM context?▼

Lossless distillation is the process of identifying and extracting all discrete facts and relationships from source documents to produce a hyper-compact distillate that preserves original information fidelity for downstream AI tasks.

Can I split a large distillate to fit different downstream context windows?▼

Yes, the distillation process supports semantically split distillates, allowing you to load specific sections as context to fit varying downstream LLM context window requirements.

How do I validate the provenance of extracted LLM distillates?▼

You validate distillates using annotated frontmatter that includes source origins, downstream consumer information, and token estimates, enabling round-trip validation and full traceability of the extracted information.

What document types work best for LLM token compression and extraction?▼

Document distillation applies to a range of source documents including briefs, discovery notes, reports, and architecture docs, extracting facts and relationships to minimize token overhead.

Do I need any external dependencies to distill documents for AI workflows?▼

No external dependencies are required to distill documents; you simply provide source documents and an optional downstream consumer parameter to generate the lossless distillate output.