agent-md-refactor

Refactors monolithic agent instruction files into a root file and linked Markdown documents.

Updated Mar 16, 2026

Search Tags:#documentation #markdown #organization #refactor #progressive disclosure #agent instructions

Ready to run

npx skills add https://github.com/Likas07/t3code-skills --skill agent-md-refactor-likas07

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: agent-md-refactor
Source: https://github.com/Likas07/t3code-skills/tree/main/skills/agent-md-refactor
Command: npx skills add https://github.com/Likas07/t3code-skills --skill agent-md-refactor-likas07

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill tackles the issue of overly long and disorganized agent instruction files (like AGENTS.md or CLAUDE.md) by applying progressive disclosure principles, making them easier to read, maintain, and use.

Core Features & Use Cases

Progressive Disclosure: Splits monolithic instruction files into a concise root file and linked, categorized sub-files.
Conflict Resolution: Identifies and helps resolve contradictory instructions.
Use Case: You have a 500-line CLAUDE.md file. This Skill will extract the essential, universal instructions into the root CLAUDE.md and move language-specific, testing, or style guidelines into separate, linked files like typescript.md or testing.md.

Quick Start

Use the agent-md-refactor skill to refactor my CLAUDE.md file.

Frequently Asked Questions about agent-md-refactor

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

FAQPage Schema

How do I organize messy agent instructions in a large markdown file?▼

To organize messy agent instructions, you refactor the monolithic file by applying progressive disclosure. The Skill extracts essential, universal instructions into a concise root file and moves detailed guidelines into linked, categorized sub-documents for better readability and maintenance.

What is progressive disclosure for agent instruction files?▼

Progressive disclosure for agent instruction files is a structuring technique that splits a monolithic document into a concise root file and linked, categorized sub-files. This separates universal rules from detailed guidelines, making the instructions easier to read, maintain, and use.

How do I refactor a CLAUDE.md file that has conflicting instructions?▼

To refactor a CLAUDE.md file with conflicting instructions, the Skill identifies and helps resolve contradictions during the reorganization. It separates rules into categorized sub-files and flags redundant or vague instructions for deletion to ensure consistency.

Can I split language-specific guidelines out of a root AGENTS.md file?▼

Yes, you can split language-specific guidelines out of a root AGENTS.md file. The Skill extracts universal instructions into the root file and categorizes language-specific, testing, or style guidelines into separate, linked markdown documents like typescript.md or testing.md.

What is the best way to clean up redundant rules in agent instructions?▼

The best way to clean up redundant rules in agent instructions is to refactor the file using progressive disclosure. The Skill flags redundant and vague instructions for deletion while organizing the remaining content into a structured root file and linked sub-documents.