How ruslanmv/bot-mmorpg-ai works
BOT-MMORPG-AI is a desktop-oriented game-learning bot: users record gameplay, train a model, select it in ModelHub, and run autonomous play. The graph covers that workflow plus setup, configuration, controller/input capture, diagnostics, and background jobs. Several relationships are documented rather than directly shown in sampled implementation; those edges cite the README or the relevant caller. Unsampled source is represented only where the README or SOURCE INDEX establishes the capability.
Main entry points: launcher.py Launcher UI · tauri.py ModelHub API · train_model.py Model training · collect_data.py Data collector
Components 23
Desktop experience
- Launcher UI
launcher/launcher.py - Setup wizard
launcher/wizard/wizard_controller.py - Desktop shell
src-tauri/src/main.rs - Game configuration
src/bot_mmorpg/config/settings_manager.py
Gameplay capture
- Dataset files
modelhub/session_manager.py - Input recorder
frontend/input_record/recorded_macro.py - Video recorder
frontend/video_record/recording.py - Data collector
src/bot_mmorpg/scripts/collect_data.py
Model learning
- Model training
src/bot_mmorpg/scripts/train_model.py - Curriculum training
src/bot_mmorpg/training/curriculum.py - PyTorch models
src/bot_mmorpg/scripts/models_pytorch.py
Bot runtime
- Bot command handlers
src/bot_mmorpg/bridge/handlers.py - Inference engine
src/bot_mmorpg/inference/engine.py - Game screen
src/bot_mmorpg/scripts/grabscreen.py
ModelHub operations
- ModelHub API
modelhub/tauri.py - Session manager
modelhub/session_manager.py - Model registry
modelhub/registry.py - Local model store
modelhub/local_store.py - Diagnostics
modelhub/diagnostics/routes.py - Background jobs
modelhub/jobs/routes.py - Job runner
modelhub/jobs/runner.py
Other components
- Player
- Game
Connections 29
- Player → Launcher UI: uses
README.md - Launcher UI → Setup wizard: opens
- Launcher UI → Session manager: manages sessions
launcher/launcher.py - Player → Input recorder: records play
README.md - Player → Video recorder: records play
- Input recorder → Data collector: captures actions
- Video recorder → Data collector: captures frames
- Data collector → Dataset files: saves dataset
README.md - Launcher UI → Data collector: starts collection
launcher/launcher.py - Launcher UI → Model training: starts training
launcher/launcher.py - Dataset files → Model training: provides data
README.md - Model training → Curriculum training: trains with
- Model training → PyTorch models: creates model
- Model training → Local model store: saves model
- Launcher UI → Model registry: selects model
launcher/launcher.py - Launcher UI → Game configuration: configures game
- Player → Bot command handlers: starts bot
README.md - Bot command handlers → Inference engine: runs inference
src/bot_mmorpg/bridge/handlers.py - Inference engine → PyTorch models: uses model
src/bot_mmorpg/inference/engine.py - Inference engine → Game screen: reads frames
- Game screen → Game: captures display
- Bot command handlers → Game: sends actions
- Desktop shell → ModelHub API: calls API
modelhub/tauri.py - ModelHub API → Session manager: manages sessions
modelhub/tauri.py - ModelHub API → Model registry: uses registry
modelhub/tauri.py - ModelHub API → Local model store: scans models
modelhub/tauri.py - ModelHub API → Diagnostics: mounts routes
modelhub/tauri.py - ModelHub API → Background jobs: mounts routes
modelhub/tauri.py - Background jobs → Job runner: submits jobs
modelhub/jobs/routes.py