AI & me

AIME Now Supports AGENTS.md - The Standard Format for AI Coding Agents

October 19, 2025

We’re excited to announce that AIME Directory now supports AGENTS.md - an open-source format that’s rapidly becoming the standard way to provide context and instructions to AI coding agents.

What is AGENTS.md?

Think of AGENTS.md as a README for AI agents. While README.md files are designed for humans - with quick starts, project descriptions, and contribution guidelines - AGENTS.md complements this by containing the extra, sometimes detailed context coding agents need: build steps, tests, conventions, and project-specific guidance that might clutter a README.

The format is simple: it’s just Markdown. No proprietary formats, no complex configuration. But it’s powerful because it’s designed specifically for how AI agents work with code.

Why Does This Matter?

AGENTS.md is already used by over 20,000 open-source projects and is supported by a growing ecosystem of AI coding tools:

  • OpenAI Codex - General-purpose CLI tooling for AI coding agents
  • Amp - AI-powered development environment
  • Jules from Google - Google’s AI coding assistant
  • Gemini CLI - Google’s command-line AI tool
  • Phoenix, RooCode, Factory, and many more

When you add an AGENTS.md file to your project, all of these tools can automatically discover and use your project-specific instructions. One file, many agents.

How AIME Directory Helps

We’ve curated 10 ready-to-use AGENTS.md files for popular frameworks and tools:

  • pnpm Monorepo - Complete guide for pnpm workspace projects
  • Next.js App Router - Instructions for modern Next.js development
  • React + Vite - Frontend development with Vite bundler
  • Python FastAPI - Backend API development with Python
  • Node.js Express - Express.js REST API patterns
  • Astro SSG - Static site generation with Astro
  • Vue 3 Composition API - Modern Vue.js development
  • Django REST Framework - Python Django APIs
  • Go Gin Framework - Go REST API development
  • Docker Compose - Multi-container application orchestration

Each file includes:

  • Setup commands for getting started
  • Code style guidelines and best practices
  • Testing instructions
  • Common issues and solutions
  • PR and commit conventions

Using AGENTS.md in Your Projects

Option 1: Browse and Download

  1. Visit aime.directory/agents
  2. Find the framework you’re using
  3. Click to view the full content
  4. Copy it to your project as AGENTS.md

Option 2: Add to a Collection

AIME’s collection feature makes it even easier:

  1. Browse agents and click “Add to Collection”
  2. Add other resources (MCPs, instructions, prompts)
  3. Click “Export ZIP”
  4. Extract and you’ll find AGENTS.md at the root level

When you export a collection, the AGENTS.md file is automatically placed at your project root - exactly where AI agents expect to find it.

Note: Per the AGENTS.md spec, a collection can only contain exactly one AGENTS.md file. If you add multiple agents to your collection, only the first one will be exported. This ensures your project has clear, focused guidance.

What Goes in AGENTS.md?

The format is flexible, but common sections include:

  • Setup Commands - How to install dependencies and run the project
  • Code Style - Project-specific conventions and patterns
  • Testing Instructions - How to run tests and what coverage is expected
  • Build Process - Compilation, bundling, and deployment steps
  • Common Issues - Gotchas and solutions specific to your project
  • PR Guidelines - Commit message format, review process

Here’s a minimal example:

# Development Instructions

## Setup

- Install dependencies: `pnpm install`
- Start dev server: `pnpm dev`
- Build: `pnpm build`

## Testing

- Run all tests: `pnpm test`
- Run specific test: `pnpm test -t "test name"`
- Coverage must be > 80%

## Code Style

- Use TypeScript strict mode
- Prefer functional components
- Single quotes, no semicolons
- Run `pnpm lint` before committing

## PR Instructions

- Format: `[package-name] Brief description`
- Always run `pnpm lint` and `pnpm test`
- Update tests for any code changes

Large Monorepos?

For large monorepos, you can use nested AGENTS.md files. Place one in each package or subdirectory. AI agents automatically read the nearest file in the directory tree, so each subproject can have tailored instructions.

For example, the main OpenAI repository has 88 AGENTS.md files - one for the root and others for specific packages.

Get Started Today

Head over to aime.directory/agents to explore our curated AGENTS.md files. Whether you’re working with Next.js, Python, Go, or Docker, we’ve got you covered.

And if you’re already using AIME Directory collections, your next export will automatically include any agents you’ve added - ready to drop into your project.

Want to Learn More?

AGENTS.md is simple, open, and already widely adopted. We’re excited to make it even easier to use through AIME Directory.

Happy coding! 🚀