Dynamic Agent Generation for Self-Adaptive Root Cause Analysis

This artifact provides the necessary code, scripts, and data to reproduce the main experimental results of the paper Dynamic Agent Generation for Self-Adaptive Root Cause Analysis accepted to SEAMS 2026. It includes the implementation of the Parser and Executor agents, configuration files for all evaluated settings, and preprocessed observability data derived from the Nezha and OpenRCA benchmarks. The goal of this artifact is to ensure that the reported results are transparent and verifiable, supporting the claims for the “Artifact Available”, “Artifact Functional”, and “Artifact Reusable” badges.