Skip to content
Need Help? Sign in
  • There are no suggestions because the search field is empty.

How to Build a Litigation Brief Bank Using the Trellis Connector | Trellis Support

A step-by-step workflow for comparing pleadings side by side, using Trellis's trial court brief bank inside your AI platform.

Overview

Most law firm brief banks only contain what that firm has filed. That's a small slice of the pleadings filed on any given claim. The Trellis Connector brings Trellis's state trial court data, including rulings and the pleadings tied to them, directly into AI platforms like Claude or ChatGPT, letting attorneys search that larger brief bank and compare pleadings side by side without switching platforms.

This article walks through a four-step workflow for finding two comparable pleadings with different outcomes and identifying the drafting choices that separated them. The example involves fraud complaints challenged by demurrer under the pleading standard from Lazar v. Superior Court, but the same workflow applies to any claim, motion type, or legal standard covered by Trellis.

For a full case study applying this workflow, see the Trellis blog: Side by side: lessons from a litigation brief bank.

Don't have the connector enabled yet? Click here to connect Trellis to Claude before starting the workflow below.

🔎 To get started with the connector, see Trellis Law MCP Connector.
🔎 For privacy information, see Privacy & Confidentiality When Using the Trellis MCP Connector.

What You Need

  • A Trellis account
  • The Trellis Connector enabled in your AI platform (this walkthrough uses Claude)
  • A claim, motion type, or pleading standard you want to research
  • Optional: a specific judge, if you want to hold that variable constant

 

Step 1: Establish the Universe of Rulings

Start broad. Ask your AI platform to pull rulings on the motion type and claim you're researching, and tell it to use Trellis.

Example prompt:

How have judges in Los Angeles County Superior Court ruled on demurrers to fraud claims? Use Trellis.

This surfaces how often the claim survives the motion across the county, giving you a sense of how large and searchable the underlying pool of rulings and pleadings is.

The Trellis Connector surfacing the pool of fraud demurrer rulings available to search.

🔎 For tips on searching motions and pleadings directly in Smart Search, see Researching Motions Filed in State Court and/or Finding Motion Templates on Smart Search.

Step 2: Hold the Key Variables Constant

A brief bank is only useful if the comparison is fair. Two complaints are only meaningfully comparable if they involve the same claim, the same pleading standard, and, ideally, the same judge. Ask your AI platform to narrow the pool accordingly.

Example prompt:

Find fraud demurrer rulings decided by [Judge Name] of the Los Angeles County Superior Court applying the pleading standard from Lazar v. Superior Court. Use Trellis.

Holding these variables constant isolates the one thing you actually want to study: the drafting choices in the complaint itself.

🔎 For analytics on how a specific judge rules on motions, see Judge Research.

Step 3: Pull the Underlying Pleadings Tied to Contrasting Outcomes

Once you've narrowed to a single judge and standard, look for two rulings that reached different results. Because Trellis links rulings to the pleadings the court evaluated, you can go straight from the ruling to the complaint behind it.

Example prompt:

Find two fraud complaints ruled on by [Judge Name] that reached different outcomes on demurrer. Show me the underlying complaints filed in each case. Use Trellis.

This is the step that turns Trellis into a brief bank rather than just a ruling search. You now have two real pleadings, decided by the same judge under the same standard, with opposite results.

A ruling explaining why one complaint's fraud allegation lacked the specificity required under Lazar v. Superior Court.

Step 4: Compare the Drafting Choices Side by Side

With both complaints in hand, ask your AI platform to isolate the specific element or paragraph that decided the outcome, rather than comparing the pleadings as a whole.

Example prompt:

Compare how each complaint pleaded the misrepresentation element of fraud. What specific drafting choices distinguish the complaint that survived demurrer from the one that didn't?

The answer typically isn't a difference in legal theory. It's a difference in specificity: whether the complaint anchors its allegations to an identifiable communication (who said what, to whom, when, and how) or refers generally to a course of dealing. Seeing both versions side by side makes that distinction concrete in a way that reading the case law alone doesn't.

The contrasting complaint, pleading the same element with the specificity the court required.

Tips for Getting Better Results

  • Hold the judge, claim, and legal standard constant. This isolates the drafting choices as the variable under comparison, rather than differences in judge, claim, or applicable law.
  • Ask for a specific element, not the whole pleading. Comparisons are sharper when you focus your AI platform on the paragraph or element the ruling actually turned on.
  • Cite the controlling case or rule. Naming the pleading standard (for example, Lazar v. Superior Court) tells your AI platform which framework to apply when it evaluates the complaints.
  • Request a side-by-side format. A table or chart makes it easier to scan two pleadings for the language that differs.
  • Read the ruling alongside the pleading. The ruling explains why one drafting choice worked and the other didn't. The pleading shows you what that choice looked like on the page.

Related Articles