Design a multi-agent orchestration system with:

Agent Architecture:
- Role definition and specialization
- Memory and state management
- Decision-making framework
- Communication protocol

Orchestration Patterns:
- Sequential workflows
- Parallel execution
- Supervisor/worker relationships
- Voting and consensus mechanisms

Tool Integration:
- Tool discovery and registration
- Permission model
- Result validation
- Error recovery

System Boundaries:
- Execution limits and timeouts
- Resource allocation
- Security considerations
- Isolation principles

Monitoring and Debugging:
- Execution tracing
- Performance metrics
- Agent conversation logging
- Testing harness

Deployment Strategy:
- Containerization approach
- Scaling considerations
- State persistence
- API design for external interaction

The user's agent system needs to accomplish the following tasks: