Managing Multiple Agents
Independent brains, a shared body
Each agent owns its own memory, personality and cognitive state, while sharing the cerebellum (plan orchestration) and the brainstem (execution engine). That yields:
- Cognitive autonomy — every agent reasons and decides independently
- Resource reuse — orchestration and execution are shared, lowering system overhead
- Parallel operation — several agents run their own tasks at the same time
- Hot reload — a single agent can restart without disturbing the others
- Lazy loading — agents are created on demand, saving resources
Four collaboration modes
| Mode | How it works | Typical use |
|---|
| Sequential | Agents process one after another in a pipeline | Draft → review → publish |
| Parallel | Several agents handle different subtasks at once | Analysing many dimensions simultaneously |
| Master-slave | One lead agent directs several subordinate agents | Project management: assigning and consolidating work |
| Consensus | Multiple agents judge independently and then vote | Risky decisions, multi-perspective verification |
Feature inventory
| Module | Status | Description |
|---|
| Project management | Implemented | Project rule configuration, multiple rules, team association |
| Team management | Implemented | Agent responsibilities, working hours, contact details |
| NeurFlow workflows | Implemented | LLM-generated workflows, node configuration, drag-and-drop editor, debugging walker |
| Group chat | Implemented | @-mentions across agents, information-storm protection, SSE real-time push |
| Task board | Implemented | Draggable cards, custom columns, live updates |
Managing agents
# add an agent
neurova agent add --agent-id <id> --name <name>
# show details / delete
neurova agent show <agent_id>
neurova agent delete <agent_id>
```