Lawve AI Becomes Anthropic Connector Partner

❤️ Lawve AI (formerly Lawvable) is now an official Anthropic Connector partner. The connector is live in Claude's Connectors directory today, part of the Claude for Legal launch. What this unlocks: Claude users can search and browse community built legal skills directly in conversation. Not templates. Not generic output. Skills from practitioners who actually do the work, with attribution, methodology, and coverage across EU AI Act compliance, Data Protection checks, specialized cross border frameworks, language specific drafting, the list goes on and on! This matters because general purpose AI in legal work has a real problem: it's good at average and bad at specialized. Subspecialty practices, jurisdictional specifics, cross regulatory analysis, domain working methods, the work that doesn't generalize. The critical point: these skills are starting points for professional review, not substitutes for legal judgment. The value is surfacing the right working method quickly, not replacing the lawyer who applies...

Updated Seven Skills on Lawve AI

🎤 Quick PSA: I've updated my skills on Lawve AI, now there are seven little helpers. If you've looked at any of them before, they're sharper now. If you haven't, the Red Team Verifier is the place to start in my opinion. The idea: AI generated legal content should face adversarial scrutiny before it reaches anyone who matters. Systematic fact checking, source validation, quality control. Not because AI gets everything wrong, but because "it looked right" is not a defensible standard in legal work ;) The Red Team skill in particular has received great feedback and interesting follow up requests and the current version is better for it. A special thank you to Marc Frey, whose input helped shaped the 2.0 build. One completely new addition I want to highlight: the Legal Test Builder. It generates fully interactive assessments as self contained HTML artifacts, complete with a countdown timer, annotated contract review tasks, model answers, and a marking rubric. Useful for testing your current skills, preparing fo...

Legal Red Team Verifier Skill

What comes after the prompt? Making legal knowledge accessible is the first step. Making it verifiable is the next. That's why I've published my Legal Red Team Verifier on Lawvable, a systematic skill for fact checking AI generated legal content. Adversarial by design. Free to use. Democratizing law doesn't just mean giving everyone access to legal reasoning. It also means giving everyone the tools to know when to trust it. As always, thank you to Antoine Louis and Malik Taiar for getting Lawve AI up and running and please check out the excellent work others have been contributing. Most recently, Oliver Schmidt Prietz has published some very high quality skills on data protection that I'd encourage any data protection professional to look at closely.

Democratizing Legal AI One Skill at a Time

Democratizing (Legal) AI – One "Skill" at a Time :) I’ve been using Anthropic’s “Skills” feature for Claude for quite some time now and have already customized my setup with 15 skills tailored for my daily work. Recently, I came across Lawvable, an open hub created by Antoine Louis for storing, sharing and discovering such “Skills.” Here's what makes "Skills" different: "Skills" are portable AI instructions that teach AIs like Claude how to perform specialized tasks. Think NDA review frameworks, GDPR compliance checklists or contract analysis methodologies. What's powerful: Anthropic released "Skills" as an open standard and OpenAI quietly adopted the same format for ChatGPT and Codex CLI. This means "Skills" are truly portable across AI platforms (fingers crossed that others will adopt the standard as well, because "Skills" are awesome). I've currently adapted and contributed three "Skills" to the platform: Vendor Due Diligence Framework: systematic assessment for IT service providers and third party vendors...