Law Firm Malpractice Insurance Comparer

The skill works from policy documents — typically PDF — and is at its best when given the full policy package for each carrier under review: the declarations page, the base policy form, every endorsement attached to that policy, and any quote letter or binder. Endorsements matter enormously; a comparison built from base forms alone is misleading, and the skill refuses to gloss over them.

Beyond the policy documents themselves, the skill benefits from:

The firm's insurance application(s), when available — these get cross-checked against firm-profile answers to surface divergences that could affect coverage.

Any written communications from the broker or underwriter that touch on coverage questions — quote summaries, email confirmations, written records of verbal representations. These get integrated alongside the policy text, with mismatches flagged.

Honest answers to a firm-profile intake at the front of the workflow. The skill needs to know what the firm actually does — practice areas, jurisdictional footprint, client types, prior-acts posture, claim history exposure, structural setup — because that is what makes the analysis specific to this firm rather than a generic comparison. A shorter intake is offered on request, with a clear warning that the shorter form degrades the analysis.

The skill performs less well when given fragmentary inputs — a declarations page without the base form, a base form without endorsements, scanned policies of poor image quality. It handles these by being explicit in the output about what couldn't be located or read, rather than guessing.

The deliverable is a coordinated pair of files. A Word executive memo is the entry point — narrative, readable, organized so a non-lawyer can follow it, ending with a list of items flagged for counsel's review. An Excel workbook sits behind it as the working document for the reviewing attorney — a row-by-row comparison matrix, supporting sheets for endorsements, applications, broker communications, and firm-profile inputs, plus methodology and disclaimer material.

Every substantive statement in either file carries a pin cite back to the specific policy language it rests on — carrier, policy number, document (base form or specific endorsement), page, section, paragraph, and a pull-quote of the operative text. Where the policy is silent on something the reference categories cover, the output says so explicitly rather than inferring. Every populated row carries confidence labels on several independent axes, so the reviewing attorney can see at a glance how much weight to put on each line.

Flags throughout the output are always framed as items for counsel to review rather than conclusions. Mandatory disclaimers appear in fixed locations. The Word memo and Excel workbook are linked by a shared ID convention so the attorney can move fluidly between the narrative and the supporting matrix.

What you will not get from this skill: 

a ranking, a recommendation, a prediction about whether a claim would be covered, or a judgment that one policy is broader, narrower, better, or worse than another. Those are calls for the attorney reviewing the package — the skill's contribution is making that review fast, organized, and grounded in cites.