Five drop-in AI skills that take an MVA matter from intake to demand — intake summary, records requests, medical chronology, case evaluation, and the demand letter. No code, no API keys, no setup. Download a skill, upload it once, and it works every time.
They're built to chain: the output of one becomes the input of the next, with a human reviewing every judgment call. Click any stage to open it.
Pick a skill and run it on a sample. This is a canned demo with synthetic facts — the real thing runs inside your AI assistant, on your file.
Every skill here is a single instruction file. There's nothing to wire up — no keys, no scripts, no subscriptions to a separate tool. You download it, upload it to your assistant once, and from then on it triggers automatically when you describe the task.
The skills draft; you decide. Each one keeps the attorney in the loop, refuses to invent facts, and flags every assumption for review before anything goes out the door.
How we keep it safe →Download the .skill file, upload it, done. No API keys or fill-in.
Skills draft and flag assumptions. You review and sign. Nothing is fabricated.
Pick a skill to read exactly what it does and see its full SKILL.md — the same file you'll download. Each is a self-contained instruction package; the demand builder also ships a fill-in letter template.
↓ Download all five skills (.zip)
The skills are deliberately scoped so each does one job and hands off cleanly to the next. You don't have to use all five — but together they cover the spine of an MVA matter.
AI is fast and fluent, and fluency is not accuracy. A confident, well-formatted answer can still contain a fabricated case, a wrong number, or a diagnosis that isn't in the record. Every skill here is built around that reality — and so should your habits be.
Treat every AI output as a lead, never a source. The signature on the demand is yours, not the model's.
OPEN and the skills will not assert them. The method, the four substantive defects v0.2.0 fixed, and the eleven errors we made finding them are all published at /verification-protocol. A library that tells you to check its work should show you its own.No technical setup. If you can upload a file, you can install one of these.
From the Skill Library, download the skill you want. You'll get a .skill file (or the SKILL.md inside it) — a plain instruction package, nothing to configure.
In Claude, go to Settings → Capabilities → Skills. (In Claude Code, drop the folder into ~/.claude/skills/.) This is a one-time step per skill.
.skill fileChoose the file you downloaded. The skill is now installed. There are no API keys, no accounts to link, and no fields to fill in.
Skills load at the start of a conversation. Say something like "draft a demand for this case" or "build a medical chronology from these records" and the matching skill triggers automatically.
The skill drafts and flags its assumptions in an "items for attorney review" block. You verify the facts and figures, make it yours, and sign. That last step is never the model's job.
TopMVA is a focused library of drop-in AI skills for motor-vehicle-accident and personal-injury practice. It covers the repetitive, high-volume drafting that eats a plaintiff's-side week — intake summaries, records requests, medical chronologies, case evaluations, and demand letters — so your team spends its time on judgment, not boilerplate.
The skills are deliberately no-code. There's nothing to integrate and nothing to maintain — each is a single instruction file you upload once. They're tuned to Missouri norms by default and built to keep a human in the loop on every judgment call.
These skills are drafting aids for licensed attorneys and their staff. They are not legal advice, they do not create an attorney-client relationship, and they do not replace attorney review. Statutes of limitation, liens, and demand terms carry real consequences — confirm everything.
Skills and materials: CC BY 4.0. Adapt and reuse with attribution.
Grab the full set as individual .skill bundles, or browse them one at a time in the library.