Show HN: AI SRE Arena, an Open Benchmark for AI SRE Agents on Kubernetes

Category: devtools

Tags: kubernetes, sre, benchmark, ai-agents, incident-response

Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 8, Utility: 7)

Project Arena is a vendor-neutral benchmark for evaluating AI SRE agents on Kubernetes. It deploys a disposable Kubernetes cluster, injects faults across 21 scenarios, captures investigation records from any product, and scores them with a configurable judge. It enables reproducible comparisons between AI-driven incident investigation tools.

Target audience: DevOps engineers, SRE teams, AI/ML researchers, and platform engineers

Repository: https://github.com/edgedelta/project-arena · Python · MIT · 12 stars

View on Hacker News