GraphJin: Governed AI agent framework with zero unsafe effects

dosco · x · 2026-08-22

GraphJin is an AI agent infrastructure based on GraphQL that provides unified governance over databases, APIs, and files. The project claims to have achieved 'zero unsafe effects' across 1,695 public benchmark attempts, capable of executing queries and operations while refusing unauthorized requests via policy checks. It has been deployed at scale in large Silicon Valley companies.

Original post →

More from coding & agent

coding & agent channel →