agents-md

Structure AGENTS.md files with reference files and failure modes.

1|Updated Mar 17, 2026
One-click install
npx skills add https://github.com/daedalus/skills --skill agents-md-daedalus
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agents-md
Source: https://github.com/daedalus/skills/tree/main/skills/agents-md-improve
Command: npx skills add https://github.com/daedalus/skills --skill agents-md-daedalus

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AGENTS.md files often suffer from inconsistent guidance and documentation sprawl across a codebase. This skill provides a structured approach to writing, auditing, and improving agent instruction files to maintain clarity and governance.

Core Features & Use Cases

  • Offer a concise entry point for agent instructions and support reference files that dive into patterns, failure modes, and best practices.
  • Enable auditing workflows to reduce over-exploration and keep documentation aligned with the codebase.
  • Help teams structure agent guidance for CLAUDE.md, COPILOT.md, and related instruction files across modules and monorepos.

Quick Start

Create a concise AGENTS.md that defines the agent instruction file scope and root references to guide downstream patterns.

Frequently Asked Questions about agents-md

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

FAQPage Schema
How do I structure AGENTS.md files to prevent guidance drift in a monorepo?

Prevent guidance drift in AGENTS.md by applying a pattern-based framework with a concise entry point and root references, delegating detailed patterns and failure modes to reference files, which keeps architecture prose bounded and aligned across codebases.

What is the best way to audit agent instruction files like CLAUDE.md and COPILOT.md?

The best way to audit agent instruction files like CLAUDE.md and COPILOT.md is to apply a structured pattern-based framework that checks for bounded scope, concise architecture prose, and proper reference file usage, reducing over-exploration and keeping documentation aligned with the codebase.

When do I need reference files for agent instructions?

You need reference files for agent instructions when your AGENTS.md guidance grows complex enough to cause over-exploration or documentation sprawl. Reference files dive into patterns, failure modes, and best practices, keeping the main instruction file concise while maintaining comprehensive governance.

Does this approach work for both small modules and large monorepos?

Yes, this approach works for both small modules and large monorepos. The pattern-based framework scales by defining a concise entry point and bounded scope in AGENTS.md, using reference files to manage guidance complexity regardless of codebase size.

Why does my AGENTS.md documentation suffer from over-exploration and how do I fix it?

AGENTS.md documentation suffers from over-exploration when guidance lacks bounded scope and structured reference files. Fix it by refactoring with a pattern-based framework that enforces concise architecture prose, establishes a clear entry point, and delegates detailed patterns to separate reference files.

How to refactor agent instruction documents to reduce documentation sprawl?

Refactor agent instruction documents by auditing existing guidance, enforcing bounded scope, and restructuring into a concise entry point with reference files. This pattern-based approach reduces documentation sprawl and aligns agent instructions with the actual codebase architecture.