This page lists three turns, the messages you send to the AI one after another in the same chat, and then a check of the answer. Keep it open next to your chat. Each turn shows the next thing to paste, with the values you entered already filled in. Nothing is sent from this page, and your conversation stays in the AI tool you are using.
Who they are, their role, and how they are aligned.
Your one-sentence theory. Every question should earn its way back to this.
The documents tied to this witness and anything they have already said under oath or in writing.
The forum and anything that governs scope, duration, or objections.
What the transcript is for. This changes the sequence and how hard you push.
How long you actually have, on the record.
Watch for these throughout
Time on the record is the real constraint. A plan listing forty topics is a plan that ends at topic twelve. Cut before you walk in, not at hour six.
The model will produce exhibit numbers and page-line cites that look right and are not. Check every record cite against your actual exhibit list and transcripts.
30(b)(6) notices, protective orders, and per-deposition time limits differ by court and by case. Confirm scope and duration before you rely on a topic being fair game.
A privilege instruction can strand a whole block. Carry a backup path for each so you are not improvising when the lock-in line is blocked.
Deposition outlines are work product. Do not paste privileged assessments or client-confidential documents unless your firm's AI policy and the engagement terms permit it.