Open Agent Harness

English · 한국어

open-agent-harness (oah)

Purpose: a harness that keeps adapting

oah helps you build an AI development workflow that fits you and keeps changing as AI improves.
It is a working reference and a guide you can adapt. A harness is the set of tools, rules and records
around an AI coding assistant.

Your AI takes care of the whole job of building and fixing a program like an expert: checking what a change
might affect, keeping the project wiki, proving the result and running the review cycle. With the recommended,
optional TypeSafe Jev judge on, it also judges difficulty and picks the model automatically through that difficulty
level in v0.3.0. You ask, answer when the AI stops for a decision, and confirm before anything lands.

Take the parts that help, or start with the whole setup and make it yours. Older rules and skills
(instructions an AI can load) can hold a newer model back or make its results worse. Keep what helps,
remove what no longer does, and change the rules as your tools change. oah keeps evolving too.
Its defaults should be easy to understand and change; a rule that gets in your way is a design problem to fix.

Why Claude implements and Codex reviews

This split comes from the author's own experience, not a measured comparison. In my use, Claude is good
at filling in details a beginner might miss while building the first version. Codex is good at checking
and correcting work within a fixed scope. So Claude writes first; Codex reviews, corrects and reviews again.

Use it with any AI setup

The oah command drives Claude Code and Codex CLI today. The ideas work with one AI or several,
including other tools and harnesses. Take the parts you need and give each stage to the AI that suits you.

What oah does for you

You describe the result you…