← back to ideas

Sentient Sentinel

8.1
ai profitable added: Tuesday July 2026 03:35

A platform for automated and continuous safety testing of open-weight AI models, providing a verifiable safety score before release, addressing Anthropic's proposal for mandatory safety testing.

180h
mvp estimate
8.1
viability grade
5
views

technology stack

Python Difficult PostgreSQL

inspired by

Anthropic wants tests, not bans