bmad-distillator

Compress source documents into lossless distillates with token estimates.

Updated Mar 24, 2026
One-click install
npx skills add https://github.com/imonmi/INTER-EDU --skill bmad-distillator-imonmi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-distillator
Source: https://github.com/imonmi/INTER-EDU/tree/main/_bmad/core/skills/bmad-distillator
Command: npx skills add https://github.com/imonmi/INTER-EDU --skill bmad-distillator-imonmi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Lossless distillation of source documents into a single dense artifact suitable for downstream LLM workflows, preserving all factual content while removing human-centric overhead.

Core Features & Use Cases

  • Produce a single distillate or semantically partitioned distillates that preserve every fact, decision, constraint, and relationship from the sources.
  • Supports an end-to-end activation: analyze inputs, compress groups, verify completeness, and optionally run a round-trip validation for high-stakes content.
  • Outputs are portable and token-estimated, enabling downstream workflows like PRD creation, architecture design, or long-form content synthesis.

Quick Start

Provide one or more source documents, optionally specify downstream_consumer and token_budget, and run the tool to generate a distillate adjacent to the primary source.

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 factual information?

Lossless distillation compresses source documents by removing human-centric overhead while preserving every fact, decision, and relationship. This produces a dense, LLM-ready artifact with token estimates for downstream workflows like PRD creation or architecture design.

What is the best way to package large documents into a single context for LLM workflows?

Packaging large documents into a single context involves analyzing inputs, compressing groups, and verifying completeness. You can generate a portable distillate or semantically partitioned parts, optionally specifying a downstream consumer and token budget to fit your pipeline.

How do I split long documents semantically for downstream LLM pipelines?

Semantic splitting for downstream LLM pipelines is achieved through fan-out routing, which partitions a distillate into semantically coherent parts. This preserves all factual content and constraints while fitting specified token budgets for tasks like long-form content synthesis.

Can I validate that document compression preserved all original facts for high-stakes content?

Yes, you can validate document compression by running an optional round-trip validation step. This high-stakes verification follows the analyze, compress, and verify activation steps to ensure the distillate is completely lossless.

Does lossless document distillation require any external dependencies or libraries?

No external dependencies are required to perform lossless document distillation. The tool operates self-contained using its internal scripts to analyze inputs, compress groups, verify completeness, and output portable distillates with token estimates.

When should I avoid using lossless distillation for information extraction?

You should avoid lossless distillation if your downstream LLM workflow requires preserving the original document's narrative structure or human-centric formatting, as the process specifically strips human overhead to produce a dense, factual artifact.