In plan mode, the plan should stay visible in the context in terminal

1) have CLI form a plan -- possibly a long and detailed plan after careful investigation
2) either accept the plan, or offer more info (last option) to recreate plan.

-->
The CLI continues as appropriate, but the plan text vanishes from the terminal scroll back buffer. I have Ctrl-t on, so I should see everything -- and most of what I do see (internal thinking) is far less interesting than the plan. I want to see the plan because

1) I can refer to it while its working without finding and opening the correct plan.md
2) it gives me confidence that it hasn't lost that valuable analysis - and hasn't lost my hopefully valuable commentary that I rejected the plan with
3) I can compare a second iteration of the plan it offers against the last one it presented.

I do not mind whether I have to have Ctrl-t on to see it, but I think it should not need that on, becuase this is "user focused" text.

What I don't know -- whether this is being hidden because it causes confusion to the agent because if its in the context? What I do know is the UX is not as good for it being hidden.

GitHub Copilot CLI 1.0.20 / Windows Terminal / inside Agency.

Reproduziere den plan mode in Windows Terminal, zuerst mit aktiviertem Ctrl-t und anschließend mit deaktiviertem Ctrl-t, und beobachte, was im terminal scrollback verbleibt, während die CLI fortfährt. Verfolge die Einstiegspunkte für die Plandarstellung und die Kontextverarbeitung; fertig ist die Aufgabe, wenn der Plan im Terminal verfügbar bleibt, während die Arbeit fortgesetzt wird, und aufeinanderfolgende Planiterationen verglichen werden können.

Vom Indexierungsmodell aus dem Issue-Text verfasst.