Prompt Paws: A Cozy 60 FPS AI Agent Bakery Game Teaching RAG, Guardrails & Context Caching

Описание и применение

Prompt Paws – это 60 FPS‑игра‑симулятор, обучающая разработчиков принципам работы LLM‑систем, включая защиту, кэширование, RAG, вызов инструментов и многопоточность.

Задача

Engineering teams need an engaging, interactive way to train developers on production LLM architecture, guardrails, caching, RAG, tool calling, and multi‑agent workflows.

Результат

Prompt Paws provides an interactive sandbox that visualizes key AI engineering concepts, enabling teams to practice guardrails, caching, RAG, tool calling, and multi‑agent workflows in a playful environment.

Рабочий процесс

• Launch the browser game at https://robinamirbahar.github.io/cozy-ai-agent-cafe-game/
• Select a 7‑shift level to start a curriculum
• Place and upgrade Robot‑Cat Chef stations on the bakery ribbon
• Observe how each station processes orders according to its AI pattern
• Use the Auto‑Chef Blueprint button to auto‑deploy optimal pipelines
• Deploy the stateless container to Google Cloud Run with one click

Доказательства

Prompt Paws: The Cozy AI Agent Cafe & Bakery – A cute 60FPS kawaii simulation game teaching RAG, Guardrails, Context Caching, Tool Calling & Multi-Agent Workflows! @google-deepmind

Источник

Play Live in Your Browser (60 FPS HTML5 Canvas): https://robinamirbahar.github.io/cozy-ai-agent-cafe-game/

Источник