do-agents-md

Write concise AGENTS.md files that give coding agents repository context and commands.

2|Updated May 25, 2026
One-click install
npx skills add https://github.com/edheltzel/Do-Skills --skill do-agents-md-edheltzel
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: do-agents-md
Source: https://github.com/edheltzel/Do-Skills/tree/main/skills/core/do-agents-md
Command: npx skills add https://github.com/edheltzel/Do-Skills --skill do-agents-md-edheltzel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Coding agents often fail in repositories because they lack clear instructions on how to build, test, lint, and navigate the codebase. Bloated or stale instruction files waste context and mislead agents, while missing guidance leads to violated architecture boundaries and broken conventions. ## Core Features & Use Cases - Structured AGENTS.md authoring: Provides a complete template covering commands, repository structure, architecture boundaries, coding standards, testing, commit conventions, and guardrails. - Map-not-manual philosophy: Keeps AGENTS.md under ~150 lines as a table of contents that points to deeper documentation instead of duplicating it. - Monorepo pattern support: Explains how to layer root and nested AGENTS.md files so package-specific rules override global ones. - Use Case: When setting up a new repository for AI coding agents, use this Skill to generate an AGENTS.md that lets any agent build, test, and contribute correctly from day one. ## Quick Start Ask the agent to write an AGENTS.md file for this repository following the do-agents-md guidelines.

Frequently Asked Questions about do-agents-md

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

FAQPage Schema
How do I write an AGENTS.md file for my repository?▼

Structure AGENTS.md as a short table of contents with sections for commands, repository structure, architecture boundaries, coding standards, testing, commit conventions, and guardrails. Keep it under roughly 150 lines and point to deeper docs rather than duplicating them.

What should I include in an AGENTS.md file?▼

Include concrete build, test, lint, and format commands, a one-line-per-directory repository map, architecture dependency rules, only non-obvious project-specific coding standards, testing patterns, commit conventions, and explicit never-do boundaries.

How long should an AGENTS.md file be?▼

Keep AGENTS.md around 100-150 lines. Longer files crowd out the actual task context, cause agents to pattern-match locally instead of navigating, and become stale quickly. Point to deeper documentation for details.

How do AGENTS.md files work in a monorepo?▼

Place a root AGENTS.md for global rules like commit conventions and shared tooling, then add nested AGENTS.md files per package for package-specific instructions. The file closest to the edited code wins, and nested files should only cover what differs from the root.

What are common AGENTS.md mistakes to avoid?▼

Avoid writing 500+ line prose documents, listing rules the linter already enforces, vague instructions like write clean code, stale conventions, and omitting the commands needed to build and test. Stale rules are worse than no rules because agents follow them faithfully.