Clarifying Questions in Agent Chat
The Agent now checks before it acts. Rather than taking a short prompt like "draft a brief" and guessing at length, audience, jurisdiction, or who you act for, it asks you a short set of questions first; and only the ones it can't already answer from your prompt or your matter.
A clear, fully specified request gets no questions at all, and once you've answered something, you won't be asked again. Nothing is assumed without telling you: if you'd rather skip the questions, the AI proceeds and lists its assumptions at the top of the output.
Benefits:
Reduces rework by confirming the format, audience, jurisdiction, and client perspective before drafting begins
Eliminates surprise charges by showing you the cost and asking for your confirmation before any InfoTrack search runs
Saves time by never asking questions your prompt has already answered
Builds trust by stating openly what the AI doesn't know, instead of papering over
What it does for the user:
Asks only the questions relevant to your task, drafting questions for documents, jurisdiction and depth for research; all in one go
Waits for your answers before drafting, researching, or running any search; for complex matters, it shows you its proposed approach first so you can approve or redirect it
Offers legal research or an InfoTrack search only where it would genuinely improve the result,
with the cost shown before anything runs
Remembers your answers for the matter: such as who you act for; so you're never asked twice
Proceeds on request if you say "just draft it", flagging every assumption at the top of the output
Step by Step Guide:
Read the clarification question.
Select the best option.
Repeat for any additional questions.
Type a response in "Something else" if needed.
Click Submit.
Review your generated output.