Reproduce “GALA: Graph-Augmented LLM Agents for Root Cause Analysis and Incident Response in Microservices” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yifang Tian, Yaming Liu, Zichun Chong, Zihang Huang, Yiran Li, Hans-Arno Jacobsen
GALA+ combines metrics, logs, and traces to rank faulty pods and guide agents through bounded investigations of candidates and their service-graph neighbors. Across two controlled benchmarks of injected faults, it achieves AC@1 scores of 74.44% and 73.33%, versus 48.35% and 47.78% for Flow-of-Action, respectively. These results concern predefined fault types, not concurrent production failures.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.