A non-coding PM had Claude Code write a Telegram RPG bot, deploy it over SSH, and run tens of thousands of test fights to rebalance classes.
A non-coding PM had Claude Code write a Telegram RPG bot, deploy it over SSH, and run tens of thousands of test fights to rebalance classes.
A project manager who doesn't code owned the design and every balance decision. Claude Code wrote all of the bot, from combat to a player market. It deployed over SSH, checked the bot was running, and built analytics that the two read together. For the class rebalance, it ran tens of thousands of test fights and proposed safe fixes.
Two evenings produced a prototype, and a live game followed in six weeks. After seven weeks it had 164 players, 109 of them active in the last day. The author ran no marketing, and paid ads did poorly.
The article gives no cost figures or details of the test-fight setup. Its stated lesson is that the human supplies the vision and the balance calls, while the AI does the execution.
Build a Telegram RPG bot where players pick from [number] classes like [class names] and fight [enemy types]. Write a simulation that runs [10,000+] test fights to check class balance and adjust stats until no class wins more than [percent]% of the time. Before deploying to my server at [host] over SSH, pause and show me the deployment steps and bot token handling for approval.
Claude Code got an old DirectX game running on modern hardware, decompiled much of it, built a headless parallel runner, and now writes full level guides with screenshots.
Plan, execute and review a cross-cutting refactor in small batches, with Claude Code making the edits and GitHub Copilot reviewing each PR.
Claude Code reads your coverage report, picks the riskiest untested modules and writes unit tests that pass, one small PR at a time.