@inproceedings{zhu-etal-2026-agentdebugx,
title = "{A}gent{D}ebug{X}: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in {LLM} Agents",
author = "Zhu, Kunlun and
Ye, Xuyan and
Han, Zhiguang and
Zhao, Yuchen and
Li, Bingxuan and
Zhang, Weijia and
Tian, Muxin and
Tang, Xiangru and
Lu, Pan and
Zou, James and
You, Jiaxuan and
Ji, Heng",
editor = "Agrawal, Sweta and
Abad, Alberto",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-demos.52/",
pages = "537--546",
ISBN = "979-8-89176-461-3",
abstract = "LLM agent failures are difficult to debug because the step where an error surfaces is often not the one that caused it. Existing observability tools replay execution traces but provide little support for identifying the root cause or translating diagnosis into recovery. We present AgentDebugX, an open-source debugging framework that organizes debugging as a closed loop of Detect, Attribute, Recover, and Rerun. At its core, DeepDebug performs multi-turn root-cause diagnosis through global trajectory understanding, structure-guided investigation, and cross-examination. On the Who When benchmark, DeepDebug achieves the best strict attribution accuracy among the evaluated methods on both tested open-weight backbones, reaching 28.8{\%} exact agent-and-step accuracy on qwen3.5-9b, compared with 21.7{\%} for the strongest single-pass baseline. On GAIA, DeepDebug repairs 13 of 73 failed tasks in a single rerun, compared with 4{--}6 for three decoupled self-correction baselines, improving overall accuracy from 55.8{\%} to 63.6{\%}. AgentDebugX exposes this workflow through a Python library, CLI, web console, and installable agentic skill. It also provides an opt-in Error Hub for sharing scrubbed failure{--}diagnosis{--}repair bundles and reusing them as debugging memory."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhu-etal-2026-agentdebugx">
<titleInfo>
<title>AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kunlun</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuyan</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhiguang</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuchen</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bingxuan</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weijia</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Muxin</namePart>
<namePart type="family">Tian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangru</namePart>
<namePart type="family">Tang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pan</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">James</namePart>
<namePart type="family">Zou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiaxuan</namePart>
<namePart type="family">You</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heng</namePart>
<namePart type="family">Ji</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sweta</namePart>
<namePart type="family">Agrawal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alberto</namePart>
<namePart type="family">Abad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-461-3</identifier>
</relatedItem>
<abstract>LLM agent failures are difficult to debug because the step where an error surfaces is often not the one that caused it. Existing observability tools replay execution traces but provide little support for identifying the root cause or translating diagnosis into recovery. We present AgentDebugX, an open-source debugging framework that organizes debugging as a closed loop of Detect, Attribute, Recover, and Rerun. At its core, DeepDebug performs multi-turn root-cause diagnosis through global trajectory understanding, structure-guided investigation, and cross-examination. On the Who When benchmark, DeepDebug achieves the best strict attribution accuracy among the evaluated methods on both tested open-weight backbones, reaching 28.8% exact agent-and-step accuracy on qwen3.5-9b, compared with 21.7% for the strongest single-pass baseline. On GAIA, DeepDebug repairs 13 of 73 failed tasks in a single rerun, compared with 4–6 for three decoupled self-correction baselines, improving overall accuracy from 55.8% to 63.6%. AgentDebugX exposes this workflow through a Python library, CLI, web console, and installable agentic skill. It also provides an opt-in Error Hub for sharing scrubbed failure–diagnosis–repair bundles and reusing them as debugging memory.</abstract>
<identifier type="citekey">zhu-etal-2026-agentdebugx</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-demos.52/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>537</start>
<end>546</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents
%A Zhu, Kunlun
%A Ye, Xuyan
%A Han, Zhiguang
%A Zhao, Yuchen
%A Li, Bingxuan
%A Zhang, Weijia
%A Tian, Muxin
%A Tang, Xiangru
%A Lu, Pan
%A Zou, James
%A You, Jiaxuan
%A Ji, Heng
%Y Agrawal, Sweta
%Y Abad, Alberto
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-461-3
%F zhu-etal-2026-agentdebugx
%X LLM agent failures are difficult to debug because the step where an error surfaces is often not the one that caused it. Existing observability tools replay execution traces but provide little support for identifying the root cause or translating diagnosis into recovery. We present AgentDebugX, an open-source debugging framework that organizes debugging as a closed loop of Detect, Attribute, Recover, and Rerun. At its core, DeepDebug performs multi-turn root-cause diagnosis through global trajectory understanding, structure-guided investigation, and cross-examination. On the Who When benchmark, DeepDebug achieves the best strict attribution accuracy among the evaluated methods on both tested open-weight backbones, reaching 28.8% exact agent-and-step accuracy on qwen3.5-9b, compared with 21.7% for the strongest single-pass baseline. On GAIA, DeepDebug repairs 13 of 73 failed tasks in a single rerun, compared with 4–6 for three decoupled self-correction baselines, improving overall accuracy from 55.8% to 63.6%. AgentDebugX exposes this workflow through a Python library, CLI, web console, and installable agentic skill. It also provides an opt-in Error Hub for sharing scrubbed failure–diagnosis–repair bundles and reusing them as debugging memory.
%U https://aclanthology.org/2026.emnlp-demos.52/
%P 537-546
Markdown (Informal)
[AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents](https://aclanthology.org/2026.emnlp-demos.52/) (Zhu et al., EMNLP 2026)
ACL
- Kunlun Zhu, Xuyan Ye, Zhiguang Han, Yuchen Zhao, Bingxuan Li, Weijia Zhang, Muxin Tian, Xiangru Tang, Pan Lu, James Zou, Jiaxuan You, and Heng Ji. 2026. AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 537–546, Budapest, Hungary. Association for Computational Linguistics.