AgentTrace: Causal Graph Tracing for Root Cause Analysis in Deployed Multi-Agent Systems
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915894594109440 |
|---|---|
| author | Wang, Zhaohui Geoffrey |
| author_facet | Wang, Zhaohui Geoffrey |
| contents | As multi-agent AI systems are increasingly deployed in real-world settings - from automated customer support to DevOps remediation - failures become harder to diagnose due to cascading effects, hidden dependencies, and long execution traces. We present AgentTrace, a lightweight causal tracing framework for post-hoc failure diagnosis in deployed multi-agent workflows. AgentTrace reconstructs causal graphs from execution logs, traces backward from error manifestations, and ranks candidate root causes using interpretable structural and positional signals - without requiring LLM inference at debugging time. Across a diverse benchmark of multi-agent failure scenarios designed to reflect common deployment patterns, AgentTrace localizes root causes with high accuracy and sub-second latency, significantly outperforming both heuristic and LLM-based baselines. Our results suggest that causal tracing provides a practical foundation for improving the reliability and trustworthiness of agentic systems in the wild. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_14688 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AgentTrace: Causal Graph Tracing for Root Cause Analysis in Deployed Multi-Agent Systems Wang, Zhaohui Geoffrey Machine Learning Artificial Intelligence Software Engineering As multi-agent AI systems are increasingly deployed in real-world settings - from automated customer support to DevOps remediation - failures become harder to diagnose due to cascading effects, hidden dependencies, and long execution traces. We present AgentTrace, a lightweight causal tracing framework for post-hoc failure diagnosis in deployed multi-agent workflows. AgentTrace reconstructs causal graphs from execution logs, traces backward from error manifestations, and ranks candidate root causes using interpretable structural and positional signals - without requiring LLM inference at debugging time. Across a diverse benchmark of multi-agent failure scenarios designed to reflect common deployment patterns, AgentTrace localizes root causes with high accuracy and sub-second latency, significantly outperforming both heuristic and LLM-based baselines. Our results suggest that causal tracing provides a practical foundation for improving the reliability and trustworthiness of agentic systems in the wild. |
| title | AgentTrace: Causal Graph Tracing for Root Cause Analysis in Deployed Multi-Agent Systems |
| topic | Machine Learning Artificial Intelligence Software Engineering |
| url | https://arxiv.org/abs/2603.14688 |