The projects we’ve published and shared — Free and Open.
A self-assessment tool for the EU AI Act, our public GitHub organization, vertical experiments and case studies co-designed with practicing professionals.
L · 01
Synthos AI Act Check
PRE-SCREENING DOSSIER · WEB APP · ITALIAN INTERFACE · EU JURISDICTION
A pre-screening dossier for European organizations, grounded in primary regulatory sources, to position AI applications across the four tiers of the EU AI Act before the August 2026 enforcement window.
For
Compliance leads, product owners, CTOs preparing for August 2026.
Inside
Six sections of primary sources (Garante Privacy, EUR-Lex, Gazzetta Ufficiale), risk classification across the four AI Act tiers, an engagement path that turns the result into operative work.
Our public organization for the code we release into the open. Public repositories, working code, experimental drops — where the methodology meets the file system.
For
Developers, technical leaders, partners who want to read what we ship.
Inside
Public repositories under permissive licenses, free to read, fork, and contribute to.
An open-source Claude Code plugin for Italian criminal-defense lawyers, built alongside practicing counsel. A vertical proof that the AI Methodology applies to regulated professional practice.
For
Italian criminal lawyers exploring AI in their daily practice. Specific to Italian procedural and substantive law.
Inside
Skills covering custody timing, statute of limitations, defensive strategy, drafting of briefs and procedural acts.
CASE STUDY · ITALIAN LEGAL LANGUAGE · PRE-REGISTERED EVALUATION
A Lab case study: can a powerful open-source Chinese model work inside a private, protected environment? We put it through Italian legal language, the hardest exam we know: 39 out of 39 citations verified against official sources, human legal review, everything published.
For
Legal professionals, innovation leads and compliance teams who want verifiable measurements of AI applied to law.
Inside
A corpus of 52 official sources frozen with cryptographic fingerprints, a multi-provider harness, a deterministic citation resolver, full logs and the complete decision record. Case study in Italian.
The methodology, taken into the public form of code.
The Lab is where the AI Methodology takes a public form. Each release maps to a principle from the Methodology Doctrine: clarity over claim, evidence over promise, vertical competence over generic approach, sovereign data, reversible engagement.
Open code is a stance: a way to let clients, peers, and counterparts read what we ship and judge it on its merits. Free tools are an obligation: when the regulation is public, the self-assessment to read it deserves to be public too.