How deestudio028-droid/jarvis works
JARVIS is a multimodal personal assistant whose main workflow runs from user input through a live assistant session, model/tool handling, and spoken or visual results. It also offers a separately runnable WhatsApp voice-call bridge, desktop and system automation, plugins, persistent memory, and a local HTTP dashboard. The README documents Gemini Live and offline/local intelligence; sampled excerpts do not show every runtime route, so some major capabilities are represented at subsystem level. Edges reflect documented statements or visible references and calls, not assumed sequencing.
Main entry points: main.py Live Assistant · live_test_bridge.py Voice Bridge Runner · whatsapp_voice_bridge.py WhatsApp Voice Bridge · llm_client.py LLM Client
Components 25
Assistant Runtime
- Live Assistant
main.py - HUD Interface
ui.py - HTTP Dashboard
dashboard/server.py
Intelligence & I/O
- LLM Client
core/llm_client.py - Speech Recognition
core/stt.py - Speech Synthesis
core/tts.py
Automation Skills
- Browser Control
actions/browser_control.py - Computer Settings
actions/computer_settings.py - Development Agent
actions/dev_agent.py - File Processing
actions/file_processor.py - File Operations
actions/file_controller.py - Web Search
actions/web_search.py - Game Updater
actions/game_updater.py - Other Assistant Skills
actions
Communications & Plugins
- Plugin Discovery
core/plugin_loader.py - Voice Bridge Runner
live_test_bridge.py - WhatsApp Voice Bridge
plugins/whatsapp_voice_bridge.py - WhatsApp Desktop Calls
plugins/whatsapp_desktop_call.py - WhatsApp Monitor
plugins/whatsapp_monitor.py - Drone Control
plugins/ky_ufo_drone.py
Configuration & State
- Conversation Memory
memory/memory_manager.py - Settings
memory/config_manager.py
Other components
- User
- Ollama Service
- WhatsApp Desktop
Connections 28
- User → Live Assistant: starts session
- Live Assistant → HUD Interface: launches interface
main.py - Live Assistant → LLM Client: requests responses
main.py - Live Assistant → Conversation Memory: loads context
main.py - Live Assistant → Browser Control: dispatches tools
main.py - Live Assistant → Computer Settings: dispatches tools
main.py - Live Assistant → Development Agent: dispatches tools
main.py - Live Assistant → File Processing: dispatches tools
main.py - Live Assistant → File Operations: dispatches tools
main.py - Live Assistant → Web Search: dispatches tools
main.py - Live Assistant → Game Updater: dispatches tools
main.py - Live Assistant → Other Assistant Skills: dispatches tools
main.py - Live Assistant → Speech Recognition: uses speech input
- Live Assistant → Speech Synthesis: uses speech output
- User → HTTP Dashboard: sends commands
dashboard/server.py - HTTP Dashboard → Live Assistant: queues commands
main.py - Voice Bridge Runner → WhatsApp Voice Bridge: runs bridge
live_test_bridge.py - Voice Bridge Runner → Speech Recognition: uses transcription
live_test_bridge.py - Voice Bridge Runner → LLM Client: requests responses
live_test_bridge.py - Voice Bridge Runner → Speech Synthesis: speaks replies
live_test_bridge.py - WhatsApp Voice Bridge → WhatsApp Desktop Calls: starts calls
plugins/whatsapp_voice_bridge.py - WhatsApp Voice Bridge → LLM Client: generates replies
plugins/whatsapp_voice_bridge.py - WhatsApp Voice Bridge → Speech Synthesis: speaks replies
plugins/whatsapp_voice_bridge.py - WhatsApp Desktop Calls → WhatsApp Desktop: controls desktop
plugins/whatsapp_desktop_call.py - Plugin Discovery → Settings: checks settings
core/plugin_loader.py - WhatsApp Monitor → Other Assistant Skills: sends messages
plugins/whatsapp_monitor.py - Ollama Service → LLM Client: serves model requests
- Live Assistant → Plugin Discovery: loads plugins
main.py