Plug-and-play AI capabilities with documented methodology, known failure modes, and ongoing validation — so you can adopt AI faster without building and testing everything in-house.
You ask. The tools produce evidence. Your assistant reasons.
Step 1
In the chat you already use. No form, no second app.
Step 2
Official text and the rulings that decided cases like yours, each dated and cited.
Step 3
It applies the rules and cites what it used.
It runs over MCP, the open standard for connecting assistants to tools. New to MCP?
One URL, no API keys, no install. Works today with Claude and ChatGPT. Free to start.
https://mcp.proofworklabs.com/mcp
Trade Compliance
HTS tariff classification with CBP ruling precedents and tariff-critical attribute extraction.
Trade Compliance
Restricted and denied-party screening against the US Consolidated Screening List, with the evidence your LLM needs to triage false positives.
Right entry among the top 10, for names written differently
3 tools · not certified
Medical Devices
FDA medical-device classification: candidate product codes, device class, the 21 CFR regulation behind them, the premarket pathway and cleared 510(k) predicate candidates.
Marketed devices whose registered product code was in the 25 candidates the find tool returns
3 tools · not certified
One certified skill, one measured figure, and a page that shows how it was measured — the scope, the interval, the cases we excluded and the failure modes we know about.
97.2% accuracy
See how we measured thisProofwork never sees your conversation and never calls a model. The tools receive the arguments of each call and nothing else.
Security & compliance details