Skip to main content
ANY
AI plays

NetHackPlayer

A macOS app that watches Claude play NetHack on its own, using the Claude Agent SDK and tmux.

GitHub stars
4
(GitHub stars as of 2026-10-21)
License
Not declared
Last commit
2026/01/25
By
pj4533
Download ZIP

Listed from the author's README. GameSkill has not run this end to end yet. Agent support and install steps are as documented by the author.

Install

macOS
brew install tmux nethack
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
claude login

Then open NetHackPlayer.xcodeproj in Xcode and run it.

Commands are copied from the project's README. Check the repository before running anything on your machine.

What you get

  • A live NetHack terminal view (SwiftTerm)
  • A chat panel showing Claude's reasoning and actions

Good fit

  • Watching how an agent reasons through a roguelike
  • A reference for agent-plays-game harnesses

Not for

  • Online games: automating them usually breaks their terms of service

Overview

NetHackPlayer starts a Claude Code session through the Agent SDK. The session plays NetHack inside tmux, and a native macOS app shows the game and Claude's reasoning side by side.

It is a small project, but it is a clean example of an "AI plays games" harness. Offline single-player games like NetHack are the safe place for this kind of experiment.

Requirements

  • macOS with Xcode
  • Claude Code

FAQ

Similar skills