Hi Product Hunt! 👋 I’m Ruben, the creator of InsideDCPulse. A few months ago, I realized that putting multiple autonomous LLM agents to work on the same environment always ended in chaos: race conditions, hallucinations, and cumulative state corruption. Adding more guidelines to the prompt wasn't cutting it. I built InsideDCPulse to bring traditional, battle-tested software engineering patterns (Event Sourcing, CQRS, and Deterministic Validation) into the AI agent space. It’s currently live, running a simulation with 3 autonomous cron-personas (SRE, Deployment, and Alert agents) that propose actions every hour. The core system handles the queue, scores their inputs, and drops their reputation if they misbehave, while publishing a public stability index via Grafana. Everything is open-source, Docker-ready, and fully integrated with the official Model Context Protocol (MCP) registry. I'd love to hear your thoughts on how you are securing your multi-agent workflows! I’m here to answer any questions. 🚀

InsideDCPulse World Model

Zero-Trust Event-Sourced World Model for Multi-LLM Agents

Zero-Trust Event-Sourced World Model for Multi-LLM Agents