Texas Hold'em practice app: 7 AI styles, friend rooms, or public tables, then fix leaks with replay, a local solver, and drills. Practice chips, no real money.
What real task does your product handle with GPT-6 Astra?
Maker
BluffKing is a browser-based Texas Hold'em practice and review app: players face seven AI opponent styles, play with friends, replay hands, and train decisions with a local solver and drills, using practice chips. We use GPT-6 Astra in Codex as an engineering agent for development and production validation. On September 9, it carried out 14 rounds of public-network load testing, analyzed latency and capacity at different action rates, published bilingual findings, and verified the production release and cleanup. This helps us turn a working multiplayer app into a measured, verifiable product. Astra's role is in the engineering workflow; poker gameplay runs on our own engine. Play: https://app.bluffking.ai. Evidence: https://bluffking.ai/en/blog/fro...
Report
Maker
📌
Hi Product Hunt! I’m building BluffKing to connect poker practice with learning from the hands you actually play. You can practice against 7 AI opponent styles, play with friends, then replay hands and work through decisions with a local solver and drills. Every table uses practice chips with no monetary value. Try it in your browser at https://app.bluffking.ai — I’d love feedback on the practice-to-review experience and which decisions you find hardest to learn.
I use GPT-6 Astra through Codex to refine the bilingual website and fix navigation issues around the training tools, then check the changes with targeted tests and real-browser verification. The workflow connected implementation, targeted testing, and browser verification across the app and its website. Poker gameplay runs in BluffKing's own engine, and strategy analysis uses our local solver; Astra is a development tool rather than the engine making poker decisions.
Hi Product Hunt! I’m building BluffKing to connect poker practice with learning from the hands you actually play. You can practice against 7 AI opponent styles, play with friends, then replay hands and work through decisions with a local solver and drills. Every table uses practice chips with no monetary value. Try it in your browser at https://app.bluffking.ai — I’d love feedback on the practice-to-review experience and which decisions you find hardest to learn.
I use GPT-6 Astra through Codex to refine the bilingual website and fix navigation issues around the training tools, then check the changes with targeted tests and real-browser verification. The workflow connected implementation, targeted testing, and browser verification across the app and its website. Poker gameplay runs in BluffKing's own engine, and strategy analysis uses our local solver; Astra is a development tool rather than the engine making poker decisions.