agent-md-refactor

Restructure bloated agent instruction files into a minimal root file and linked documentation.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/piyushmal13/kandhogaya --skill agent-md-refactor-piyushmal13
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-md-refactor
Source: https://github.com/piyushmal13/kandhogaya/tree/main/.kilocode/skills/agent-md-refactor
Command: npx skills add https://github.com/piyushmal13/kandhogaya --skill agent-md-refactor-piyushmal13

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Bloated agent instruction files like AGENTS.md, CLAUDE.md, and COPILOT.md grow into unwieldy documents that waste cognitive and context tokens, make it hard to locate relevant guidance, and increase the risk of conflicting instructions.

Core Features & Use Cases

  • Identify contradictions and essential content that must live in the root file.
  • Group remaining instructions into focused, linked documentation files (e.g., typescript.md, testing.md, code-style.md, git-workflow.md, architecture.md).
  • Create a minimal root file plus a scalable, organized file hierarchy that users can navigate for different topics.

Quick Start

Refactor my AGENTS.md into progressive disclosure by creating a minimal root file and organized linked documentation.

Frequently Asked Questions about agent-md-refactor

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

FAQPage Schema
How do I refactor bloated AGENTS.md or CLAUDE.md files?

Apply progressive disclosure by extracting essential instructions into a minimal root file under 50 lines, grouping related guidance into topic-specific linked files like testing.md, and flagging vague content for deletion to create a scalable documentation hierarchy.

What is progressive disclosure for AI agent instruction files?

Progressive disclosure for agent instruction files means structuring documentation as a minimal root file linking to self-contained, topic-specific files. This hierarchy prevents unwieldy documents, saves cognitive tokens, and makes relevant guidance easier to locate.

How do I organize agent instructions into linked documentation files?

Identify core instructions for the root file, then group related rules into focused linked files like typescript.md, testing.md, code-style.md, git-workflow.md, and architecture.md. This creates a navigable, scalable file hierarchy for different topics.

Does agent-md-refactor work with COPILOT.md files?

Yes, this refactoring approach works with AGENTS.md, CLAUDE.md, COPILOT.md, and similar agent instruction files. It restructures any bloated agent documentation into a minimal root file and organized linked documentation hierarchy.

What is the best way to clean up redundant content in agent documentation?

Clean up redundant agent documentation by identifying contradictions, flagging vague content for deletion, and extracting essentials into a root file. Group remaining valid instructions into self-contained linked files to eliminate conflicting guidance.