Exploring Next
Exploring Next — Ep 228 w/ Justy & Cody — Reddit - The heart of the internet
Lundrog built an open-source framework called agent-guardrails-template to control AI coding agents and prevent them from breaking codebases. The system uses four safety laws, active enforcement via a Go MCP server, and risk-based decision matrices to reduce AI-caused incidents by 78%.