Publication
SRE-Marathon: A Continuous, Change-Driven Benchmark for Autonomous Site Reliability Agents
Yifang Tian / Yingjian Bai / Yifeng He / Zichun Chong / Yuanchen Gao / Yiran Li / Hans-Arno Jacobsen
arXiv, 2026 ยท ArXiv Preprint
Abstract
Read the abstract and full preprint at the source.
Research outcomes
SRE-Marathon project website and leaderboard
Project website and paper-reported leaderboard for continuous, change-driven SRE evaluation. Its GitHub repository contains the website source, not the benchmark implementation.
Additional notes
This preprint studies continuous SRE operation with overlapping incidents in a live Kubernetes environment. Its evaluation follows the progression from correlation and localization to repair using recorded system evidence.