One question across five regimes is really five questions, and the shortcut everyone takes is to research the first one properly and then go hunting for its features in the other four. It feels efficient. What it actually does is quietly reshape four legal systems to look like the one you happened to start with.

The output looks fine, too. That's the problem. A neat comparison table is exactly as convincing when the columns don't mean the same thing.

Research each jurisdiction on its own terms

So each jurisdiction is worked on its own terms, and deliberately kept apart while that happens. Nothing from the German answer colors the French one, because they are not compared until each stands on its own.

Keeping them separate is what prevents the specific failure this work is prone to. It is a discipline about sequence rather than a claim about speed.

Where a concept exists in one regime and simply has no counterpart in another, you get told that instead of getting an approximate equivalent invented to keep the row tidy.

Want to try it?

Ask your Strawberry companion: “Compare how this question is treated across these jurisdictions. Keep the columns like-for-like and tell me where they converge and diverge.”

Skill

Compare a question across jurisdictions

Compare this question across these jurisdictions.

Make the columns mean the same thing

Normalisation is where a comparison is won or lost, and most of the traps look like nothing until someone acts on them:

  • Thresholds measured on different bases. Headcount in one regime, turnover in another, and they are not interchangeable.
  • Obligations that fall on differently defined actors, so "the provider" means two different companies.
  • Penalties written as a fixed cap in one place and a percentage of global turnover in another.
  • Dates where one column means "in force" and the next means "applicable from".

Where a difference won't normalise, it stays visible in the table rather than getting flattened into a cell that reads cleanly and misleads.

What the table is actually for

The table is not the deliverable. What you actually need is where the regimes converge, where they genuinely diverge, and what someone operating across all of them has to do on Monday.

Taking the strictest position on each dimension is one possible baseline where a single policy has to satisfy every regime at once, but it is not the legal answer and it does not hold everywhere. Jurisdiction-specific exceptions stay visible, and where two regimes genuinely conflict, that conflict is reported rather than resolved by picking the stricter one.

Every cell is a citation

It's worth saying plainly: a comparison table is a dense stack of legal claims wearing a spreadsheet. Each entry has to point at a provision that exists, says what the cell claims, is current, and belongs to the jurisdiction it sits under.

Misattribution is both most likely and most damaging in exactly this format, because a row is short enough to look self-evidently right. Anything unverified is flagged rather than left looking as solid as the cells around it.

When you compare the same regimes regularly, the accepted jurisdictions, dimensions, and sources can be saved so the next comparison starts from your shape rather than a blank table.

How to brief a new regulation with AIWhen one instrument deserves a briefing of its own. How to map a market with AIThe same discipline applied to a commercial landscape rather than a legal one.
Official Strawberry skill
Copy skill file Download skill file
name
compare-across-jurisdictions
description
Compare how a legal question is treated across several jurisdictions in one structured table, showing where regimes converge, where they diverge, and what that means for an entity operating across them. Use for multi-market compliance, expansion, and cross-border structuring questions.

Compare Across Jurisdictions

Answer the same legal question in several jurisdictions and show what the differences actually mean. The value is in the comparison being genuinely like-for-like, which is also the hardest part.

1. Fix the question and the jurisdictions

Agree the precise legal question and the exact list of jurisdictions. Confirm what the comparison is for — market entry, a compliance program, structuring, or a policy decision — because it decides which dimensions matter.

Ask whether the user needs the position as it stands today, or as it will stand once instruments already adopted become applicable. Those produce different tables.

2. Research each jurisdiction on its own terms

Research each jurisdiction separately and completely before comparing. The failure mode here is reading the first jurisdiction thoroughly and then looking only for its features in the others, which quietly imports one regime's structure onto another that does not share it.

Use the local primary law, and local secondary sources to check your reading. Where a concept exists in one regime and simply has no counterpart in another, say that rather than forcing an equivalent.

Keep client and party details out of search queries throughout.

3. Normalise before tabulating

Make sure each column means the same thing across rows. Thresholds measured differently, obligations falling on differently-defined actors, penalties expressed as caps versus percentages, and dates meaning "in force" versus "applicable from" all produce tables that look comparable and are not.

Where a difference cannot be normalised, keep it visible in the table rather than flattening it into a tidy cell.

4. Build the comparison

One row per jurisdiction, with columns chosen for the decision — typically applicable law, key provisions, scope and thresholds, obligations, enforcement and penalties, and current status.

Then write the part that makes it useful: where the regimes converge, where they genuinely diverge, and what an entity operating across all of them has to do in practice. Where a single policy must satisfy every regime at once, the strictest position on each dimension is one possible baseline. Offer it as that rather than as the legal answer, keep jurisdiction-specific exceptions visible, and report an irreconcilable conflict as a conflict rather than resolving it by picking the stricter rule.

5. Verify every cell

Each entry is a citation. Confirm the provision exists, says what the cell claims, is current, and belongs to the jurisdiction it sits under. Cross-jurisdictional work is where misattribution is both most likely and most damaging. Flag any cell that could not be verified rather than leaving it to look as solid as the rest.

6. Deliver and continue

Lead with the practical implication, then the table, then the caveats.

Use strawberry/legal/research-a-legal-question when one jurisdiction turns out to need real depth, and strawberry/legal/brief-a-regulation when a single instrument deserves its own briefing.

Preserve the accepted jurisdictions, dimensions, and sources as a custom skill when the user compares the same regimes repeatedly.