Skip to content

Home

The Home screen is where you see everything at once — every agent’s status, every issue’s position in the pipeline, and the controls that determine how work flows.

The stepper runs across the top of the Home screen. Each step maps to a status column on your board: To Do → In Progress → AI Review → Human Review → Done. Issue counts update in real time as agents claim work, create PRs, and pass reviews.

Click any step to filter the view to issues in that status.

Autopilot controls whether Cate dispatches work to agents automatically.

When autopilot is on (the default), Cate continuously matches idle agent tabs to the highest-priority issue in To Do. As soon as an agent finishes one task, it picks up the next. Reviews work the same way — when a PR reaches AI Review, an idle reviewer claims it.

When autopilot is off, agents sit idle until you explicitly assign work. Use this when you want to control exactly which issue an agent picks up, or when you need to pause all agent activity — during a deploy freeze, while you reorder the board, or when you want to review the plan before agents start.

Toggle autopilot from the stepper. The setting applies to the current project and persists across sessions.

Each agent appears as a card below the stepper. Cards show the agent’s current state: idle, working on an issue, or reviewing a PR. Click a card to open the agent’s terminal and watch it work in real time.

Use the menu on an agent group to bulk rename agents or adjust the number of worker and reviewer slots.

The Home screen is designed to be glanceable. A healthy board has issues flowing left to right through the stepper. If issues pile up in one column, that’s your signal:

  • Stacking in To Do — agents may be stuck or you need more worker slots
  • Stacking in AI Review — add a reviewer slot so reviews keep pace with coding
  • Stacking in Human Review — that’s your queue; time to review and merge