How lokilogesh001/optimus works
OPTIMUS is a desktop personal AI assistant whose principal workflow is user command submission through the Control Center, orchestration into plans and delegated work, tool-mediated execution, and verification of outcomes. The repository also documents voice interaction, memory, workflows, model routing, and specialized agents; these remain distinct branches rather than being forced into one pipeline. Paths and edges are grounded in the README and SOURCE INDEX. Several subsystems are represented from documented capabilities despite unsampled implementations, so their internal links are omitted where evidence is unavailable.
Main entry points: orchestrator.py Central Orchestrator · tools.py Tool Registry · service_bridge.py Service Bridge · engine.py Workflow Engine
Components 23
Interaction
- Control Center
gui/optimus - Service Bridge
gui/optimus/service_bridge.py - Voice Interaction
core/interaction
Execution Coordination
- Central Orchestrator
core/orchestrator/orchestrator.py - Execution Events
core/orchestrator/events.py - Adaptive Planning
core/planning - Mission Engine
core/missions/engine.py - Workflow Engine
core/workflows/engine.py - Specialized Agents
core/agents
Intelligence
- Model Gateway
core/intelligence - Model Providers
core/intelligence/providers - Shared Memory
core/memory - Personal Intelligence
core/personal_intelligence/engine.py - Proactive Engine
core/proactive/engine.py - Self Improvement
core/self_improvement
Desktop Capabilities
- Tool Registry
core/orchestrator/tools.py - Filesystem Service
core/filesystem/service.py - Screen Vision
core/vision - Application Intelligence
core/application_intelligence
Safety and Evidence
- Agent Sandboxing
core/sandbox - Verification Engine
core/verification/engine.py - Security Policy
core/security/policy_engine.py
Other components
- User
Connections 25
- User → Control Center: submits commands
- Control Center → Service Bridge: forwards requests
- Service Bridge → Central Orchestrator: dispatches tasks
gui/optimus/service_bridge.py - Central Orchestrator → Adaptive Planning: plans goals
- Central Orchestrator → Specialized Agents: delegates subtasks
- Specialized Agents → Agent Sandboxing: executes within
- Central Orchestrator → Workflow Engine: executes workflows
core/workflows/engine.py - Workflow Engine → Central Orchestrator: dispatches steps
core/workflows/engine.py - Central Orchestrator → Tool Registry: dispatches tools
- Tool Registry → Filesystem Service: reads and writes
core/orchestrator/tools.py - Tool Registry → Screen Vision: focuses windows
core/orchestrator/tools.py - Tool Registry → Application Intelligence: routes interactions
core/orchestrator/tools.py - Central Orchestrator → Model Gateway: routes model requests
- Model Gateway → Model Providers: selects providers
- Central Orchestrator → Shared Memory: shares context
- Workflow Engine → Verification Engine: validates outcomes
core/workflows/engine.py - Central Orchestrator → Verification Engine: verifies deliverables
- Central Orchestrator → Execution Events: publishes lifecycle events
- Service Bridge → Execution Events: observes events
gui/optimus/service_bridge.py - Voice Interaction → Service Bridge: feeds interaction
gui/optimus/service_bridge.py - Tool Registry → Security Policy: checks permissions
core/orchestrator/tools.py - Mission Engine → Specialized Agents: coordinates subtasks
core/missions/engine.py - Personal Intelligence → Shared Memory: stores personal context
- Proactive Engine → Mission Engine: continues missions
core/proactive/engine.py - Self Improvement → Security Policy: applies policy
core/self_improvement/promotion_engine.py