Practical Tools: The Resolution Tracker & AI Prompt Architect
I did not see 2026 starting with unauthorized military action on foreign soil. Apparently, we’re all living a poorly written Jack Ryan fan fiction spin-off.
But amidst the chaos on the world stage, our cases continue to lurch forward. The docket doesn’t care about congressional oversight or approval.
I’m starting the new year with free tools to make your life a little easier. Not less chaotic or uncertain, but a little bit easier.
The Morello Mediation Resolution Tracker
First, two tools in one: an offer tracker and a bracket tracker, wrapped in a single Google Sheet: Morello Mediation Resolution Tracker.
The Offer Tracker records the round, the time, and the amount of the demand or offer. It then automatically calculates the spread and midpoint.
I’ve also added a visualization. You can see when you’re getting close to convergence and how quickly either side is moving. This won’t make the other side reasonable, but it will make tracking and understanding their moves a little bit easier.
The Bracket Tracker records the party proposing the bracket and the top and low end of the bracket. It then automatically calculates the spread, the implied midpoint, the spread of the midpoints, and the illustrious midpoint of the midpoints.
I do not endorse all the crazy midpoint gymnastics that some counsel use, but I’m also not denying that they exist. Tracking these numbers automatically makes the gymnastic event a little bit easier.
[Get the Morello Mediation Resolution Tracker Here]
https://docs.google.com/spreadsheets/d/14b4Oc2SOy_Iwrf7HAXbBH6rOGyMEiUED48qJs_OeJb0/copy
(Tab 1: Offer Tracker. Tab 2: Bracket Tracker.)
The Morello Mediation Prompt Architect
Second, the Morello Mediation Prompt Architect.
I’ve refined the advice on using LLMs and turned it into a practical prompting guide.
I know many of you are using LLMs (ChatGPT, Claude, Gemini) to summarize transcripts or edit briefs. I also know many of you are terrified of "hallucinations" or accidentally leaking client data.
The problem isn't the AI; it's the prompt.
If you treat an LLM like a magic 8-ball, you’ll get garbage. If you treat it like a prediction machine that needs very specific instructions to predict what you want, you may get something useful.
I’ve built an 8-step "Prompt Architect" that forces you to structure your request safely. It uses a "Sandwich Method" to ensure the AI reads your data but follows your rules.
It also includes "The Billboard Rule" for redaction and specific "Recipes" for editing briefs and stress-testing arguments.
The best part of this guide: you don’t actually need to read it. Just attach it to your preferred LLM and write:
Use the attached guide to build me a prompt to: [DESCRIBE WHAT YOU WANT THE LLM TO DO].
Your preferred LLM should return a formatted and fleshed out prompt that you can copy and paste into a new chat.
[Get the Morello Mediation Prompt Architect here]
https://docs.google.com/document/d/1skWzslSzb-ja-LgFfH4mW_sv6qL3rxVcX6zz3J3nLj8/copy
This guide won’t stop the LLM from hallucinating or turn it into a thinking machine rather than a prediction machine, but it should make using LLMs a little bit easier.
I hope everyone had a wonderful holiday season and that we are all braced appropriately for the upcoming year.
Monday x Morello Mediation.
None of this is legal advice. Your mileage may vary.