How to use Claude to build a Jungle board game
A developer used one Claude prompt to build Jungle within existing infrastructure, reported issue-free playtesting, and used a second prompt to renumber it for publication.
Original prompt
Shared by the source author. See source notes for context.
Create game 197 "Jungle" based on the Chinese board game "dou shou qi". Since it's a board game it can use a regular controller (doesn't need websockets), but please include scoring (backend authoritative/anticheat) and sound effects (ui/lib/sfx.ts)
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source poster / publisher
- pzxc
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- The author reported building the game with one prompt, finding no issues during playtesting, and using a second prompt solely to prepare its numbering for publication.
- Human involvement
- The author playtested the game and reported finding no issues before requesting the publication-related renumbering.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Prompted Claude to create Jungle within an existing game project.
- Playtested the generated game and reported finding no issues.
- Sent a second prompt to renumber the game for publication.
pzxc shared the prompt used to build Jungle, a game based on the Chinese board game dou shou qi. The author used Claude Sonnet 5.0 on Medium and requested creation of game 197 within an existing project. The prompt specified that a regular controller was suitable and that WebSockets were unnecessary for this board game.
The request also required scoring to be authoritative on the backend, with anticheat, plus sound effects through the project's existing ui/lib/sfx.ts library. These were requirements in the supplied prompt; the post did not provide an implementation audit of those features. The author explicitly noted that substantial infrastructure, including the sound-effects library, had already been built in earlier sessions.
After the initial generation, the author playtested the game and reported finding no issues. A second prompt was then used solely to renumber the game in preparation for publishing. The source does not describe further changes or a detailed test procedure.
Who it's for: Developers adding a board game to an established project with existing controllers, scoring infrastructure, and sound tools.
Prompt
Original prompt
Source: Hacker News · pzxc
Frequently asked questions
Was the game built entirely from a blank project?
No. The author said much of the infrastructure, including the sound-effects library, had been built in previous sessions.
Why was a second prompt needed?
After playtesting and finding no issues, the author used a second prompt only to renumber the game so it was ready for publishing.
More details
About this account
Claude generated the game from one implementation prompt, followed by a separate renumbering request after human playtesting.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=48849092. Prompt: Original (quoted from source). Sourced by the Agents In Use pipeline.
Credit: pzxc on Hacker News
Source published
Added to Agents In Use
Source checked
Tools and topics
Claude Sonnet 5.0 · Existing game infrastructure · ui/lib/sfx.ts
Board games · Jungle · Game development · Playtesting
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.