A real AI agent setup splits planning, execution, and review
AI development work is split across separate roles instead of being handled by one tool. An plans the work and assigns tasks, but does not change the code itself. One agent runs commands, one agent reviews what was written, and one agent researches unclear points.
also go through the execution agent, so the same part of the system is not both planning and approving the work. A live dashboard shows incoming requests, the current phase, and error counts. Cache handles many requests, while only some work goes out fresh to the model.
The setup is tested by forcing failures on purpose, including pushing a provider until it hits a , checking whether retries happen, and confirming that a fallback works when the first provider stops responding.
Key points
- The plans and delegates but does not edit code directly.
- Separate agents handle command execution, review, and research.
- go through the execution agent to keep approval separate from planning.
- A dashboard tracks requests, current phase, and error counts in real time.
- Cache handles many requests, and only part of the work is sent fresh to the model.