Free AI skills for Bahamian lawyers
Reusable Claude and ChatGPT-style skill patterns for source checks, review packs, structure notes, and counsel questions.
In plain English
- Reusable Claude and ChatGPT-style skill patterns for source checks, review packs, structure notes, and counsel questions.
- Use this as a starting point for learning the workflow, not as legal advice.
- The useful output should be clear enough for a professional to review.
5 min read
Use this as a pre-review step before a memo goes to counsel: sources, assumptions, and missing evidence stay attached to the matter.
A helpful source check has a simple job: show what is supported, what is missing, and what a lawyer or advisor should review next.
Why free skills matter
Many people will first meet legal AI through a simple assistant: Claude, ChatGPT, Codex, or another tool already on their desk. That is a good place to learn, as long as the task is narrow and the material is appropriate to use.
A skill makes that learning safer and more useful because it gives the model a defined job. Instead of asking it to answer a legal question, the skill asks it to check a draft, list missing sources, identify assumptions, and prepare questions for a lawyer to review.
The point is to help Bahamian lawyers, advisors, and students build good habits early: use AI to organize the work, keep sources visible, and make review easier.
- Use free skills with non-confidential or properly approved material only.
- Treat the output as a checklist, not legal advice.
- Keep sources and assumptions visible in the final work product.
A useful Bahamas source-checking skill
The source-checker has a modest but important job: help the reader understand whether each important statement has support. For a Bahamas matter, that usually means separating statute, SCB-administered legislation, regulator context, BFSB industry material, law-firm practice notes, filing assumptions, market practice, and open legal questions.
A good source-checking output should say: here are the propositions that need authority, here are the sources cited, here are the gaps, here are assumptions that may require counsel judgment, and here are facts that should be requested before anyone relies on the draft. If a draft says SMART Fund, ICON, DARE Act, foundation, or financial and corporate service provider, the checker should name the actual route and source.
That is helpful because it gives the lawyer a map of where the work is ready and where it still needs attention.
What to ask the model
The prompt should be explicit about the role. The model should act as a pre-review assistant for a Bahamas legal or financial services matter. It should not invent authorities. It should flag missing authorities. It should distinguish confirmed facts from assumptions. It should return a review table that can be checked against SCB, BFSB, Central Bank, and practitioner sources.
The best format is simple: issue, draft statement, source support, missing source, factual assumption, Bahamas route, risk level, and question for counsel. That lets the lawyer use the output as a review checklist.
If the model cannot identify a source from the material provided, it should say so plainly. That honesty is useful. It tells the reviewer what to check next.
- Ask for a table, not prose.
- Force the model to separate facts, law, assumptions, and questions.
- Do not let the model cite sources it cannot inspect or legal routes it cannot name.
How this becomes a repeatable workflow
The same pattern can grow into a repeatable workflow. The source-checking task connects to the memo, the document plan, the diagram, and the review queue. If a source is missing, it becomes a follow-up task. If an assumption needs judgment, it becomes a review item.
That is the bridge from learning to practice. The skill helps the model perform one useful job. A workspace keeps that job attached to the matter so the next reviewer can inspect it.