[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82562-en":3,"doc-seo-82562-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82562,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","Knowledge-Enhanced Agentic Vulnerability Repair","Frontier foundation models accelerate vulnerability discovery while reducing costs, yet remediation often cannot keep pace. Existing Automated Vulnerability Repair (AVR) methods struggle to reliably identify vulnerability root causes and underuse prior fix knowledge, weakening real-world patch generation. KEAREPAIR proposes an agentic AVR pipeline that grounds patches in verified program facts and high-level vulnerability knowledge: it extracts dual-view multi-dimensional knowledge from historical pairs, retrieves diagnostic facts via a tool-augmented ReAct agent, then performs retrieval-augmented, closed-loop validated patch refinement.","Knowledge-Enhanced Agentic Vulnerability Repair  \nSicong Cao†, Hao Ma†, Le Yu†􀀌, Kangyi Ding‡, Xiaolei Liu‡, Terry Yue Zhuo§ , Bo Wang¶ , Xingwei Lin♢ , Xiaobing Sun♠ , Linzhang Wang♣ , David Lo♡  \n†Nanjing University of Posts and Telecommunications ‡National Interdisciplinary Research Center of Engineering Physics §Alibaba Qwen ¶Beijing Jiaotong University ♢Zhejiang University ♠Yangzhou University ♣Nanjing University  \n♡ Singapore Management University  \narXiv :2607 .00820v 1 [ cs . SE] 1 Jul 2026  \nAbstract—Frontier foundation models have significantly accelerated vulnerability discovery while slashing costs, but the bigger challenge is how the remediation side keeps up. Despite recent progresses in Automated Vulnerability Repair (AVR), current solutions struggle to reliably pinpoint vulnerability causes, and insufficiently utilize the prior fix knowledge to guide the patch generation process, undermining their effectiveness in practice.  \nTo address this gap, we propose KEAREPAIR, a novel agentic AVR approach that grounds patch generation in verified program facts and high-level vulnerability knowledge. Specifically, KEAREPAIR first extracts dual-view multi-dimensional knowledge items from historical vulnerability-patch pairs, and constructs dedicated retrieval knowledge bases. It then employs a toolaugmented agent that performs ReAct-style reasoning to collect verified program facts for vulnerability diagnosis. Finally, based on the diagnostic results, KEAREPAIR performs knowledgelevel retrieval-augmented patch generation and iteratively refines patches through a closed-loop validation process. Experimental results show that KEAREPAIR significantly outperforms existing AVR approaches on 55 reproducible C/C++ vulnerabilities. When paired with Gemini-3.1-Pro, KeaRepair successfully repairs 46 vulnerabilities, achieving a repair rate of 83.64% . Moreover, KEAREPAIR fixes nine unique vulnerabilities that the state-of-theart baseline PatchAgent cannot address, and further demonstrates strong cross-language generalizability.  \nI. INTRODUCTION  \nAs the cybersecurity capabilities of frontier foundation models like Claude Mythos [1] and GPT-5.5-Cyber [2] continue to evolve, the cost and effort required to find and exploit software vulnerabilities have all dropped dramatically. As of May 2026, the public Common Vulnerabilities and Exposures (CVE) program [3] has disclosed 29,120 new vulnerabilities, an increase of approximately 39.89% year-over-year. The resulting backlog far outpaces the resources available to triage them, creating a strong demand for tools that can quickly generate correct patches for known vulnerabilities [4] .  \nExisting Efforts. Recent progresses in Large Language Models (LLMs) have sparked interest in their application to Automated Vulnerability Repair (AVR) [5], primarily due to their superior reasoning capabilities compared to traditional rule-intensive [6],[7] or popular data-driven approaches [8], [9] . Early attempts primarily follow the one-step repair paradigm, employing prompt engineering [10], reinforcement learning [11], [12], or Retrieval-Augmented Generation (RAG) [13] to directly output candidate patches. Afterwards, a series of LLM-based patching agents [14]–[16] have been proposed to autonomously localize faults and synthesize patches in an iterative workflow.  \nLimitations. While promising, existing LLM-based approaches suffer from two fundamental Limitations.  \nL1: Unreliable Root Cause Analysis. Identifying the root cause of a vulnerability and providing an optimal code location to apply patches is typically the gold standard in most AVR approaches [8], [9], [13] . However, this assumption does not align with the reality faced by developers: even the state-of-theart localization tool struggles to precisely pinpoint the exact statements that need to be fixed [17] . More importantly, root causes often do not intersect with the fix location, in fact they may be far apart [18] . Instead, agentic ","cbCairrUk1lPz2w1","https://ap.wps.com/l/cbCairrUk1lPz2w1","pdf",2013082,1,12,"English","en",105,"# Introduction\n## Existing Efforts\n## Limitations\n### Unreliable Root Cause Analysis\n### Insufficient Utilization of Vulnerability Knowledge\n## Our Work","[{\"question\":\"What problem does KEAREPAIR aim to solve in automated vulnerability repair?\",\"answer\":\"It targets two gaps: unreliable root cause analysis and insufficient utilization of vulnerability knowledge to guide patch generation.\"},{\"question\":\"How does KEAREPAIR ground patch generation in reliable information?\",\"answer\":\"It uses a tool-augmented agent to collect verified program facts for vulnerability diagnosis, and then conditions patch generation on retrieval results at the knowledge level.\"},{\"question\":\"How effective is KEAREPAIR compared with existing AVR approaches?\",\"answer\":\"Experiments on 55 reproducible C/C++ vulnerabilities show clear improvements, and with Gemini-3.1-Pro it repairs 46 vulnerabilities for an 83.64% repair rate, including nine cases the PatchAgent baseline cannot address.\"}]",1784181544,30,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"knowledge-enhanced-agentic-vulnerability-repair","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/knowledge-enhanced-agentic-vulnerability-repair/82562/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does KEAREPAIR aim to solve in automated vulnerability repair?","Question",{"text":75,"@type":76},"It targets two gaps: unreliable root cause analysis and insufficient utilization of vulnerability knowledge to guide patch generation.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does KEAREPAIR ground patch generation in reliable information?",{"text":80,"@type":76},"It uses a tool-augmented agent to collect verified program facts for vulnerability diagnosis, and then conditions patch generation on retrieval results at the knowledge level.",{"name":82,"@type":73,"acceptedAnswer":83},"How effective is KEAREPAIR compared with existing AVR approaches?",{"text":84,"@type":76},"Experiments on 55 reproducible C/C++ vulnerabilities show clear improvements, and with Gemini-3.1-Pro it repairs 46 vulnerabilities for an 83.64% repair rate, including nine cases the PatchAgent baseline cannot address.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":28,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]