dualize-docs

Convert Markdown directories into bot/ fact files and human/ explanations with insights.md.

8|2|Updated May 3, 2026
One-click install
npx skills add https://github.com/ejoongseok/claude-settings --skill dualize-docs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dualize-docs
Source: https://github.com/ejoongseok/claude-settings/tree/main/claude-code/skills/dualize-docs
Command: npx skills add https://github.com/ejoongseok/claude-settings --skill dualize-docs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

When a directory accumulates many Markdown project documents, it becomes hard to understand what is fact vs context and what insights only emerge by comparing everything together.

Core Features & Use Cases

  • Two-layer restructuring: Generates a bot/ layer for AI-only “facts” and a human/ layer for human-friendly explanations and diagrams.
  • Cross-document meta analysis: Creates human/insights.md to surface inconsistencies, hidden connections, gaps, blind spots, and forward-looking priorities based on reading all source Markdown together.
  • Optional source archiving: Supports moving original Markdown into an archive/YYYY-MM-DD/ tree when requested.
  • Deterministic, tooling-focused workflow: Uses directory validation, Markdown discovery, and strict separation rules (facts in bot/, why/how in human/, meta in insights).

Quick Start

Run the skill on a target directory to regenerate bot/ and human/ outputs and produce human/insights.md from the combined Markdown context.

Frequently Asked Questions about dualize-docs

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

FAQPage Schema
How do I split project documentation into AI facts and human-readable explanations?

To split project documentation, this Skill dualizes a directory of Markdown into a bot/ layer of AI-facing facts and a human/ layer of explanatory narratives. It applies strict separation rules to translate multiple files into two audience-appropriate layers.

What is the best way to find contradictions across a large backlog of Markdown files?

Finding contradictions across a Markdown backlog requires cross-document meta analysis. This Skill reads all source files together and generates a structured insights artifact that surfaces inconsistencies, hidden connections, gaps, and confidence-tagged findings.

How do I restructure multi-file Markdown documentation for an AI agent?

Restructuring Markdown for an AI agent requires generating a dedicated bot/ directory containing only deterministic facts. This Skill ingests comprehensive Markdown context and unifies multiple files into these separate AI-facing fact references.

Can I archive original Markdown sources while generating documentation insights?

Yes, you can archive original Markdown sources while generating insights. This Skill supports optional source archiving, moving your original files into an archive/YYYY-MM-DD/ tree when requested during the dualization workflow.

Does this documentation restructuring approach work for unifying disconnected project files?

Yes, this approach works for unifying disconnected project files by applying directory validation and Markdown discovery. It checks multiple accumulated documents for contradictions and translates them into unified bot and human layers.