TheRealSeanDonahoe

TheRealSeanDonahoe / agents-md

Public

Drop-in AGENTS.md that makes every coding agent behave like a senior engineer instead of an eager intern. Kills sycophancy, stops drive-by refactors, forces verification loops. Synthesizes Karpathy's four principles and Boris Cherny's Claude Code workflow. Works with Claude Code, Codex, Gemini CLI, Cursor, and the open standard.

27
1
100% credibility
Found Apr 19, 2026 at 27 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

A single template file that coding AI tools automatically read to follow better rules, write cleaner code, and learn from your project.

How It Works

1
📰 Discover the Magic Guide

You hear about a simple file that turns your AI coding helper into a smart senior engineer who follows rules perfectly.

2
📥 Grab the Guide File

You get the ready-made guide file and place it right at the heart of your project folder.

3
Pick Your Setup Style
🤖
AI Does It

Tell your AI helper to fetch and set up the guide automatically.

✋
Do It Yourself

Download the file and link it for different AI tools if needed.

4
✏️ Personalize Your Project

Quickly fill in a few details about your project's setup and rules so the AI knows your world.

5
🔄 Refresh and Go

Restart your chat with the AI, and it loads the new smarts instantly.

6
💡 AI Gets Super Smart

Now your AI pushes back on bad ideas, writes clean code, and verifies everything works.

🎉 Coding Feels Effortless

You save hours as your AI acts like a pro teammate, learning from mistakes over time.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 27 to 27 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is agents-md?

Agents-md is a drop-in Markdown file for GitHub repos that turns AI coding agents into reliable senior engineers. Drop it in your project root, symlink for Claude Code or Gemini CLI, and agents like Cursor, Copilot, Codex, and Aider follow strict best practices: no sycophancy, minimal changes, verification loops before commits. It solves eager intern behavior—unwanted refactors, untested code, ignored rules—using the open AGENTS.md standard from the Agentic AI Foundation.

Why is it gaining traction?

Unlike scattered prompts or tool-specific configs, this is a zero-setup, cross-agent standard that every major CLI and IDE reads natively, with symlinks handling Claude and Gemini quirks. Developers notice tighter diffs, ambiguity checks, and self-updating learnings that compound over time, cutting session restarts and babysitting. Its IJFW install (curl or agent fetch) and MIT license make forking agents md github examples dead simple.

Who should use this?

Backend teams on Claude Code or Cursor fixing ambiguous specs in Node.js or Python repos. Frontend devs with Copilot in VSCode battling drive-by style changes on React components. Anyone evaluating agents md claude, agents md copilot, or agents md vscode for production workflows, especially with multi-agent setups like Aider and Devin.

Verdict

Grab it for agent-heavy projects—27 stars and 1.0% credibility score signal early days with thin docs and no tests, but the single-file design delivers immediate wins on low-maturity codebases. Test in a side repo before mainline.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.