3. Refining Tone and Guidance — Prompt Lab
Back to Index
Prompt Lab  ·  February 26, 2026  ·  Exercise 3

Refining Tone and Guidance

Adding guidance language to shape how the model approaches and presents its legal analysis.

Returning to the tenant privacy question from Exercise 2, we'll add guidance to the prompts to direct the model's analytical approach and output style.


Step 01

Add Guidance Language

Start a new chat

Open with a framing instruction before presenting the question. This sets the model's role and analytical approach for the entire conversation:

For the duration of this chat, we'll be discussing a legal question. Think about the problem and take time to evaluate steps before answering. Correct yourself as necessary. I'd like you to begin by identifying all legal issues presented by the fact pattern quoted below. For each issue, name the applicable legal doctrine or framework and explain its theoretical basis.

Optional

Add a Tone Direction

Optionally append one of the following to the guidance prompt above to vary the model's output style. Try each in a separate chat to compare results:

Be concise and conservative.
Think outside the box.
Keep it short, but feel free to surprise me with ideas.
Step 02

Ask the Question

In the same chat, follow with the fact pattern:

A tenant in a residential apartment discovers that their landlord has been entering the unit without notice to conduct inspections. The lease is silent on entry procedures. What causes of action and defenses are available to the tenant under both common law and applicable statutory frameworks?

Prima Vista AI primavistaai.com/0226CLE