<table> <tr><td> <a href="https://www.reddit.com/r/ClaudeAI/comments/1rwyzpn/i_built_a_browser_game_where_you_fight_corporate/"> <img src="https://preview.redd.it/9t14mntmtrpg1.png?width=640&crop=smart&auto=webp&s=9b84ee7ec275a162fc39f329b7d4e7035bbda1e0" alt="I built a browser game where you fight corporate AI bots using real consumer laws - now with 36 cases" title="I built a browser game where you fight corporate AI bots using real consumer laws - now with 36 cases" /> </a> </td><td> <div class="md">What it is: 36 levels, each one a corporate or government AI that wrongly denied you something - flight refund, visa, medical authorization, gig worker deactivation. You argue back with real laws. The AI's confidence drops as you find the right arguments. New this week: after every win there's a "What you just used" panel - the law you cited, what it actually means, and how you'd use it in a real dispute. One-day build that changes the feel significantly. Stack: Vanilla JS, Node/Express, Claude Haiku as the AI engine. Each bot has a system prompt with a resistance scoring system - Claude returns <code>{message, resistance, outcome}</code> JSON on every turn and the game reads it directly. The interesting part: prompt design. Each bot has a personality, starting resistance (60–95), and specific legal arguments that reduce it by defined amounts. Main challenge was Claude breaking character on sensitive scenarios (medical denials, disability) to announce it's made by Anthropic. Fixed by framing the whole thing as an educational simulator in the system prompt. <a href="https://

I built a browser game where you fight corporate AI bots using real consumer laws - now with 36 cases

Related Skills

Design Critique Bot

Statistical Analyzer

Wiki Builder