agent-md-refactor

Refactor monolithic agent instruction files into modular linked documentation.

Updated Jul 24, 2026
One-click install
npx skills add https://github.com/imMamdouhaboammar/kaku-chatgpt-harness --skill agent-md-refactor-immamdouhaboammar
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-md-refactor
Source: https://github.com/imMamdouhaboammar/kaku-chatgpt-harness/tree/main/.agents/skills/agent-md-refactor
Command: npx skills add https://github.com/imMamdouhaboammar/kaku-chatgpt-harness --skill agent-md-refactor-immamdouhaboammar

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill resolves the issue of bloated, unmanageable agent instruction files (like CLAUDE.md or AGENTS.md) that waste context tokens and contain contradictory or vague rules.

Core Features & Use Cases

  • Progressive Disclosure: Automatically splits monolithic instruction files into a minimal root file and categorized, linked sub-files.
  • Contradiction Detection: Identifies and flags conflicting instructions before restructuring.
  • Use Case: If your CLAUDE.md has grown to 300 lines of mixed testing, style, and architecture rules, this skill will reorganize it into a clean structure with a 50-line root file and dedicated files for each topic.

Quick Start

Use the agent-md-refactor skill to analyze and split my current CLAUDE.md file into a modular documentation structure.

Frequently Asked Questions about agent-md-refactor

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

FAQPage Schema
How do I split a monolithic CLAUDE.md file into modular documentation?

Split it by applying progressive disclosure principles to separate the monolithic instruction file into a minimal root file and categorized, linked sub-files, improving context efficiency and instruction clarity.

What is progressive disclosure for agent instruction files?

Progressive disclosure for agent instruction files is the practice of restructuring bloated configurations like AGENTS.md into a minimal root file that links to categorized sub-files, optimizing context efficiency and reducing wasted tokens.

Can I detect contradictory rules in my AGENTS.md before restructuring?

Yes, contradiction detection identifies and flags conflicting instructions within your AGENTS.md before restructuring begins, ensuring logical categorization of technical guidelines and preventing contradictory or vague rules from persisting.

What is the best way to optimize context tokens wasted in project configuration files?

The best way to optimize context tokens is to refactor project configuration files into a modular documentation structure, reducing a 300-line monolithic root file to a 50-line version with dedicated files for each topic.

When do I need to refactor my agent instructions into linked sub-files?

You need to refactor agent instructions into linked sub-files when your root configuration grows to hundreds of lines of mixed rules, wasting context tokens and making it difficult to maintain logical categorization of technical guidelines.