Discover and explore top open-source AI tools and projects—updated daily.
LLM-driven coach for Hearthstone Battlegrounds
Top 97.5% on SourcePulse
DeepBattler is an LLM-powered coaching tool for Hearthstone Battlegrounds, designed to provide real-time strategic advice and voice-assisted guidance. It targets Hearthstone players aiming to improve their gameplay and climb ranks, offering insights comparable to top-tier players.
How It Works
DeepBattler integrates with the Hearthstone Deck Tracker (HDT) via a custom plugin. The plugin monitors game state, outputs data as JSON, and feeds this into a Python agent. This agent leverages LLMs (OpenAI GPT or Google Gemma) to analyze the game state and provide strategic recommendations, including voice interaction.
Quick Start & Requirements
HearthDb.dll
and HearthstoneDeckTracker.exe
as project references, setting "Copy Local" to True, and placing the compiled DeepBattlerPlugin.dll
into the HDT plugins directory. Configure the game_state.json
path in DeepBattlerPlugin.csproj
.pip install openai playsound==1.2.2
, configure API key in Openai_caller.py
, run python Openai_caller.py
.pip install keras_hub jax keras gtts playsound==1.2.2
, set environment variables (KERAS_BACKEND
, XLA_PYTHON_CLIENT_MEM_FRACTION
), prepare game_state.json
and Prompt.txt
, run python Gemma_caller.py
.Highlighted Details
Maintenance & Community
Licensing & Compatibility
Limitations & Caveats
The plugin setup requires familiarity with Visual Studio project referencing and file placement. The license explicitly prohibits commercial use. Compatibility is dependent on matching HDT and DLL versions.
6 months ago
Inactive