Attributa

Attributa is a privacy-first attribution intelligence tool designed for ethical AI development. It is hosted at alawein/attributa and operates under a P3 priority in active development.

The tool provides three core analysis capabilities. First, it detects AI-generated content by analyzing text through local ML models, specifically GLTR running via WebAssembly and DetectGPT. Second, it validates citations by checking them against CrossRef DOIs to verify that referenced sources exist and are accurately cited. Third, it scans code for CWE (Common Weakness Enumeration) security vulnerabilities, surfacing potential weaknesses in codebases under review.

A defining architectural principle of Attributa is that all primary analysis runs client-side. No data leaves the browser unless the user explicitly opts in, ensuring that sensitive text, code, and research materials remain under the user's control. This privacy-first posture makes it suitable for working with proprietary or pre-publication content.

The frontend is built with React, TypeScript, and Vite, using shadcn/ui for the component library and Zustand for state management. Supporting Python tooling appears in the declared stack for auxiliary scripts and analysis helpers that complement the browser-based layer. Attributa uses the Devkit project, which provides shared configuration packages and repo templates across alawein repositories.