zeke

zeke / agents.md

Public

My global instructions for AI agents like OpenCode

16
1
89% credibility
Found May 23, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

This repository shares a personal instruction file for AI coding assistants. The author created a set of guidelines that they load into every AI coding session to tell their helper how to behave: what to check before making changes, how to write commit messages, when to ask for clarification, and how to structure pull requests. By making this public, they invite others to copy, adapt, or simply learn from these conventions. The project also points to a community website where these ideas are documented more broadly.

How It Works

1
🔍 Discovering a smarter way to work with AI

You hear about a clever idea: giving your AI coding assistant a permanent set of instructions so it always knows how you like to work.

2
📖 Reading about agents.md

You find a public collection of instructions someone uses with their AI helpers and realize it's been shared for anyone to borrow.

3
✨ Finding your preferred style

You browse through the rules and discover helpful ideas: how to write clear commit messages, how to structure pull requests, and when your AI helper should ask questions instead of guessing.

4
📋 Picking what works for you

You select the conventions that match your workflow and start customizing your own instruction file with the parts that resonate most.

5
🤖 Your AI helper learns your preferences

You load your personalized instructions into your AI coding assistant, and it immediately understands how you like things done.

🎉 Coding together, smoothly

Your AI assistant now follows your conventions automatically, saving you time and keeping your projects consistent without constant reminders.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 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?

A personal configuration file that defines how you want AI coding assistants to behave across all your projects. Instead of repeating the same instructions every session, this file acts as a persistent system-level guide—setting rules for verification, commit style, PR templates, and even when the agent should push back on your requests. It is markdown-based and designed to live in your home directory, loading automatically into every coding session.

Why is it gaining traction?

AI coding tools are exploding in adoption, but most developers waste time re-explaining their preferences in every new conversation. This project solves that by centralizing your agent configuration once and letting it persist across sessions. The agents.md convention is starting to build momentum as a community standard, with tools like OpenCode, Claude Code, and Codex already reading files with this naming pattern. It appeals to developers who want consistent, opinionated AI assistance without constant hand-holding.

Who should use this?

Developers tired of repeating "use conventional commits" or "always write tests first" every session. Teams wanting to standardize how AI agents behave in shared development environments. Anyone maintaining dotfiles who wants their agent preferences to follow them across machines. Technical leads establishing guardrails for AI-assisted code reviews.

Verdict

Worth exploring if you live in AI-assisted workflows. The 0.8999999761581421% credibility score indicates this is a real-world, working implementation of the agents.md convention. At 16 stars it is early-stage—minimal docs, no formal examples—but the README is clear enough to get running. Fork it, steal what works, adapt it to your own setup.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.