How kevalparmar75/optimus-assistant works
Optimus is a voice-controlled, Transformers-themed desktop assistant. Its main workflow is voice command → supervisor routing → specialist agent → spoken response; the app also supports browser and screen control, general tools, memory and notes, reminders, and code generation or execution. The README documents several capabilities not visible in sampled implementation, so those links are cited to the README. The source excerpts do not show the complete LangGraph wiring or all reminder and speech behavior.
Main entry points: main.py Command supervisor · memory_agent.py Memory agent · chat_agent.py Chat agent · reminder_agent.py Reminder agent
Components 20
Voice and HUD
- Assistant runtime
main.py - Character HUD
ui/hud.py - Spoken response
main.py
Routing and agents
- Command supervisor
main.py - Chat agent
agents/chat_agent.py - Browser agent
agents/browser_agent.py - Code agent
agents/code_agent.py - Memory agent
agents/memory_agent.py - Reminder agent
agents/reminder_agent.py
Assistant capabilities
- Vision agent
agents/vision_agent.py - Browser control
agents/browser_agent.py - Screen capture and actions
agents/vision_agent.py - Code tasks
agents/code_agent.py - Conversation memory
agents/memory_agent.py - Notes
agents/memory_agent.py - Reminder scheduling
agents/reminder_agent.py
State and tools
- Agent state
state.py - Tool registry
tools/registry.py
Other components
- Voice user
- External services
Connections 27
- Voice user → Assistant runtime: gives command
README.md - Assistant runtime → Command supervisor: routes command
README.md - Command supervisor → Agent state: passes state
main.py - Command supervisor → Chat agent: dispatches to
README.md - Command supervisor → Browser agent: dispatches to
README.md - Command supervisor → Code agent: dispatches to
README.md - Command supervisor → Memory agent: dispatches to
README.md - Command supervisor → Reminder agent: dispatches to
README.md - Command supervisor → Vision agent: dispatches to
README.md - Assistant runtime → Character HUD: updates display
main.py - Browser agent → Browser control: executes actions
agents/browser_agent.py - Browser agent → Screen capture and actions: delegates vision
agents/browser_agent.py - Vision agent → Screen capture and actions: captures and acts
agents/vision_agent.py - Code agent → Tool registry: calls tools
agents/code_agent.py - Chat agent → Tool registry: calls tools
agents/chat_agent.py - Memory agent → Tool registry: wires and calls
agents/memory_agent.py - Reminder agent → Tool registry: registers tools
agents/reminder_agent.py - Memory agent → Conversation memory: stores and recalls
agents/memory_agent.py - Memory agent → Notes: reads and writes
agents/memory_agent.py - Reminder agent → Reminder scheduling: schedules jobs
agents/reminder_agent.py - Reminder agent → External services: notifies user
README.md - Chat agent → External services: queries model
agents/chat_agent.py - Code agent → External services: queries model
agents/code_agent.py - Vision agent → External services: queries vision model
agents/vision_agent.py - Assistant runtime → Spoken response: speaks reply
README.md - Assistant runtime → Memory agent: saves conversation
README.md - Chat agent → External services: searches web
README.md