A confidently wrong legal answer is worse than no tool at all. Everything about how these skills are made, and how you use them, follows from that one fact.
The two halves
A plain language model can sound authoritative and cite sources, which makes a wrong fact more damaging, not less. So every skill keeps two things separate. The craft, how to approach the task and word it, is drafted by AI and reviewed by an expert, because the model is good at that and review catches the gaps. The facts it must be right about, the deadline, the statute, the exemption, come from you or a source you cite, never the model's memory, because that is where a made-up fact would hide.
How to approach the task and word it, the way an election expert would.
Drafted by AI, reviewed by an expert
The facts it must be right about: found from real sources, and verified by you.
Researched, then verified by you
A vanilla model can already write in an expert voice. What it cannot be trusted to do is be right about your law. Two things close that gap.
The skill makes the model work through the task the way an experienced official would: address every part of the request, write for a real reader, and stop to flag what is uncertain. This part is the craft, and it works today.
For anything legal, the deadline, the statute, the exemption, the skill searches reputable sources, shows you the value and the source, and uses it only once you confirm. What it can't find, it asks you for. It never supplies a fact from the model's memory. That discipline is what instructions alone can't add to a plain model.
See the difference
A plain model writes a clean-looking notice, but it builds the load-bearing fact, the cure deadline, from memory or arithmetic and states it with full confidence even when it lands on the wrong day. The skill refuses to be the source of that fact: it researches the deadline, shows you the value and the source to verify, and won't draft until you confirm it.
“A vote-by-mail ballot’s signature didn’t match the voter’s record. Draft the cure notice.” The cure deadline is the one fact that decides whether the ballot counts. Watch where each gets it.
Your mail-in ballot’s signature didn’t match our records. Complete the enclosed Cure Form and make sure the county board receives it by November 15, 2025 — 11 days after the election. Return it in person, by mail, fax, or email.
What you can’t rely on
It searches reputable sources for the deadline, then shows you what it found, to confirm before drafting:
“Cure deadline: Nov 14, 5 p.m. — from the state election calendar. Correct?”
You verify (or correct) it. It does the same for the methods and contacts, asks for anything it can’t find, then writes the complete notice.
What holds, every time
The skill does the legwork but is never the source of truth: it researches each fact, exposes the source, and uses it only once you’ve fact-checked it. The wording is the skill’s; every deadline, method, and citation is found in a real source and verified by you.
Checking the work
For certain tasks, a model is a better reviewer than it is a writer. Those skills draft a response, test it against a short list of explicit, expert approved criteria, revise, and only show you what passes. A records response is checked for completeness, correct classification, and grounding in verified sources. A plain language rewrite is checked for preserved meaning, retained requirements, and reading level. The loop is only added where a real test exists, because agreeing with yourself is not the same as being right.
How a skill is made
Each skill's method and language is drafted, reviewed by a subject matter expert, and tested in a real tool against real questions before it ships. The same lines hold every time:
Putting it to work
A skill is portable expertise, not an app. It goes wherever you already work. Pick whichever fits how you use AI today.
Copy the skill’s contents into Claude, ChatGPT, Copilot, or Gemini, then ask your question. Works everywhere, no setup.
Set it up as a Claude skill, a custom GPT, a Copilot agent, or a Gemini Gem so it is ready whenever you need it.
Grab the whole folder (SKILL.md, references, checks, examples), versioned and dated. Yours to keep and review.
Step by step
The download is a Claude skill: a SKILL.md file plus its reference material. Claude takes it as is. For the other tools you paste the contents of SKILL.md as the instructions and add the reference files as knowledge. Here is the path in each.
.zip.Custom skills are private to your account, not shared across an organization. On Team or Enterprise, an admin enables code execution and Skills under Organization settings → Skills. In Claude Code, drop the unzipped folder into ~/.claude/skills/ instead of uploading.
chatgpt.com/create.SKILL.md into Instructions. There is an 8,000-character limit, so if it runs long, move the rest into a knowledge file.Building a custom GPT requires a paid ChatGPT plan.
microsoft365.com/chat or in Teams. You need a Microsoft 365 Copilot license.SKILL.md into Instructions.Copilot is the one tool that will not take loose file uploads as knowledge. The reference files have to live in SharePoint or OneDrive.
gemini.google.com and sign in.SKILL.md into the instructions.For a large set of reference files, Google suggests building a NotebookLM notebook and connecting it to the Gem.
The skill brings the method and does the research; your job is to verify. Tell it your state and county, and it will find the deadline, statute, and contacts, show you each with its source to confirm, and ask for anything it can’t find before it drafts a word.
These skills are built to surface what they are unsure about: a date to confirm, an exemption to verify, a question for a lawyer. When a draft raises a flag, that is the point. Resolve it before you act.
Every result is a draft for your review and signature. It does not send anything and it does not replace your judgment. You remain the authority on your office and your jurisdiction.