Currently, if no custom agent is created, one or several default agents handle tasks automatically.
These agents should be visible in the agent list in a non-editable form. If a custom agent with the same role as a default agent is created, the default agent should be disabled.
This aligns with the broader goal of reducing opacity and improving intuitiveness in the application.
At the moment, when no custom agent is defined, it is unclear who is actually handling tasks. There is a default agent prompt section in the settings, but it is not clear when these agents are triggered, whether they are actively executing something, and there is even less visibility into their logs.
Two possible approaches:
- Keep default agents, but make them visible so the user can clearly understand what is happening in their project.
- Remove default agents entirely. During project onboarding, require users to create the minimum set of agents needed to move a task from Planning to Closed. Default prompts can be suggested, but users should still be free to define their own instructions.
This would significantly reduce opacity and make it clear why a task moves or changes state, instead of it appearing to happen automatically without knowing which agent is responsible or what it is doing outside of task logs.
Currently, if no custom agent is created, one or several default agents handle tasks automatically.
These agents should be visible in the agent list in a non-editable form. If a custom agent with the same role as a default agent is created, the default agent should be disabled.
This aligns with the broader goal of reducing opacity and improving intuitiveness in the application.
At the moment, when no custom agent is defined, it is unclear who is actually handling tasks. There is a default agent prompt section in the settings, but it is not clear when these agents are triggered, whether they are actively executing something, and there is even less visibility into their logs.
Two possible approaches:
This would significantly reduce opacity and make it clear why a task moves or changes state, instead of it appearing to happen automatically without knowing which agent is responsible or what it is doing outside of task logs.