{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,11]],"date-time":"2026-07-11T18:11:26Z","timestamp":1783793486937,"version":"3.55.0"},"reference-count":37,"publisher":"Springer Science and Business Media LLC","issue":"3","license":[{"start":{"date-parts":[[2026,6,27]],"date-time":"2026-06-27T00:00:00Z","timestamp":1782518400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,6,27]],"date-time":"2026-06-27T00:00:00Z","timestamp":1782518400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Alma Mater Studiorum - Universit\u00e0 di Bologna"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Grid Computing"],"published-print":{"date-parts":[[2026,9]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    Kubernetes has become the foundation of modern cloud-native infrastructure, yet its operational complexity remains a persistent barrier. Administrators must navigate a vast API surface, manage heterogeneous workloads, and coordinate tasks across disconnected tools\u2014often requiring precise commands, declarative configuration files, and deep domain expertise. This paper presents\n                    <jats:bold>KubeIntellect<\/jats:bold>\n                    , a Large Language Model (LLM)-powered system for end-to-end Kubernetes management through natural language. KubeIntellect spans all major categories of Kubernetes operations\u2014read, write, delete, exec, access control, and lifecycle management\u2014through a supervisor-coordinated set of domain-specialized agents, with human-in-the-loop (HITL) confirmation on all mutating operations. Operations outside the static tool set are handled by the Code Generator Agent, which synthesizes, validates, and registers new Kubernetes tools at runtime. The Code Generator Agent executes synthesized tools in an in-process Python REPL rather than a separate process or container; process-level isolation between synthesized code and the host runtime is\u00a0therefore not enforced, and a defense-in-depth model comprising static analysis, API-call validation, and mandatory human-in-the-loop review constitutes the primary mitigation. Migration to pod-level isolation is a planned hardening step. Evaluation on a live four-node Kubernetes cluster (170 pods across 18 namespaces) shows: a\n                    <jats:bold>75% pass rate<\/jats:bold>\n                    (12\/16; 95% CI: 51%\u201391%; mean rubric score 31.2\/40) on a 16-scenario controlled fault-injection corpus scored on an 8-dimension LLM-judge rubric; a\n                    <jats:bold>+25 percentage-point improvement<\/jats:bold>\n                    over a tool-less GPT-4o baseline on the same scenarios (75% vs. 50%); a\n                    <jats:bold>93% query resolution rate<\/jats:bold>\n                    (186\/200) with an\n                    <jats:bold>81.8% synthesis success rate<\/jats:bold>\n                    (63\/77 novel tool requests) on a 200-query operational corpus; and end-to-end latency in the\n                    <jats:bold>7\u201310\u00a0s<\/jats:bold>\n                    range at a mean API cost of\n                    <jats:bold>$0.036\/query<\/jats:bold>\n                    for read-only workloads and\n                    <jats:bold>$0.039\/query<\/jats:bold>\n                    overall. A reproducible demo environment is available on a public managed-Kubernetes service, with a local single-node option for readers without cloud access. These results demonstrate that domain-specific multi-agent orchestration, structured HITL confirmation, and runtime tool synthesis together yield substantially higher task completion than general-purpose LLM reasoning on Kubernetes operations.\n                  <\/jats:p>","DOI":"10.1007\/s10723-026-09837-6","type":"journal-article","created":{"date-parts":[[2026,6,27]],"date-time":"2026-06-27T10:42:58Z","timestamp":1782556978000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["KubeIntellect: A Modular LLM-Orchestrated Agent Framework for End-to-End Kubernetes Management"],"prefix":"10.1007","volume":"24","author":[{"given":"Mohsen","family":"Seyedkazemi\u00a0Ardebili","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrea","family":"Bartolini","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,6,27]]},"reference":[{"key":"9837_CR1","unstructured":"Kubernetes SIGs.: kind \u2014 kubernetes IN docker (2019). https:\/\/kind.sigs.k8s.io. Accessed 2026"},{"key":"9837_CR2","unstructured":"Khan, A.F., Khan, A.A., Mohamed, A., Ali, H., Moolinti, S., Haroon, S., Tahir, U., Fazzini, M., Butt, A.R., Anwar, A.: LADs: Leveraging LLMs for AI-driven devops. arXiv:2502.20825 (2025)"},{"key":"9837_CR3","unstructured":"Malul, E., Meidan, Y., Mimran, D., Elovici, Y., Shabtai, A.: GenKubeSec: LLM-based kubernetes misconfiguration detection, localization, reasoning, and remediation. arXiv:2405.19954 (2024)"},{"key":"9837_CR4","unstructured":"Cohen, O.S., Malul, E., Meidan, Y., Mimran, D., Elovici, Y., Shabtai, A.: KubeGuard: LLM-assisted kubernetes hardening via configuration files and runtime logs analysis. arXiv:2509.04191 (2025)"},{"key":"9837_CR5","unstructured":"Kikuta, D., Ikeuchi, H., Tajiri, K.: ChaosEater: Fully automating chaos engineering with large language models. arXiv:2501.11107 (2025)"},{"key":"9837_CR6","doi-asserted-by":"crossref","unstructured":"Zhang, L., Zhai, Y., Jia, T., Huang, X., Duan, C., Li, Y.: AgentFM: Role-aware failure management for distributed databases with LLM-driven multi-agents. arXiv:2504.06614 (2025)","DOI":"10.1145\/3696630.3728492"},{"key":"9837_CR7","unstructured":"Chen, Y., Pan, J., Clark, J., Su, Y., Zheutlin, N., Bhavya, B., Arora, R., Jha, S., Xu, T., Deng, Y.: STRATUS: A multi-agent system for autonomous reliability engineering of modern clouds. In: Advances in Neural Information Processing Systems (2025)"},{"key":"9837_CR8","doi-asserted-by":"crossref","unstructured":"Parthasarathy, K., Vaidhyanathan, K., Dhar, R., Krishnamachari, V., Muhammed, B., Kakran, A., Akshathala, S., Arun, S., Dubey, S., Veerubhotla, M., Karan, A.: Engineering LLM powered multi-agent framework for autonomous CloudOps. arXiv:2501.08243 (2025)","DOI":"10.1109\/CAIN66642.2025.00031"},{"key":"9837_CR9","doi-asserted-by":"crossref","unstructured":"Chen, Y., Xie, H., Ma, M., Kang, Y., Gao, X., Shi, L., Cao, Y., Gao, X., Fan, H., Wen, M., et\u00a0al.: Automatic root cause analysis via large language models for cloud incidents. In: Proceedings of the Nineteenth European Conference on Computer Systems (EuroSys), pp. 1\u201316, (2024)","DOI":"10.1145\/3627703.3629553"},{"key":"9837_CR10","doi-asserted-by":"crossref","unstructured":"Lee, C., Yang, T., Chen, Z., Su, Y., Lyu, M.R.: Eadro: An end-to-end troubleshooting framework for microservices on multi-source data. In: Proceedings of the 45th International Conference on Software Engineering (ICSE), (2023)","DOI":"10.1109\/ICSE48619.2023.00150"},{"key":"9837_CR11","doi-asserted-by":"crossref","unstructured":"Shetty, M., Chen, Y., Somashekar, G., Ma, M., Simmhan, Y., Zhang, X., Mace, J., Vandevoorde, D., Las-Casas, P., Gupta, S.M., et\u00a0al.: Building AI agents for autonomous clouds: Challenges and design principles. In: Proceedings of the 2024 ACM Symposium on Cloud Computing, pp. 99\u2013110, (2024)","DOI":"10.1145\/3698038.3698525"},{"key":"9837_CR12","unstructured":"Chen, Y., Shetty, M., Somashekar, G., Ma, M., Simmhan, Y., Mace, J., Bansal, C., Wang, R., Rajmohan, S.: AIOpsLab: A holistic framework to evaluate AI agents for enabling autonomous clouds. arXiv:2501.06706 (2025)"},{"key":"9837_CR13","unstructured":"Liu, X., Yu, H., Zhang, H., Xu, Y., Lei, X., Lai, H., Gu, Y., Ding, H., Men, K., Yang, K., et\u00a0al.: AgentBench: Evaluating LLMs as agents. In: International Conference on Learning Representations, (2024)"},{"key":"9837_CR14","unstructured":"Jha, S., Arora, R., Watanabe, Y., Yanagawa, T., Chen, Y., Clark, J., Bhavya, B., Verma, M., Kumar, H., et\u00a0al.: ITBench: Evaluating AI agents across diverse real-world it automation tasks. arXiv:2502.05352 (2025)"},{"key":"9837_CR15","unstructured":"Xu, F.F., Song, Y., Li, B., Tang, Y., Jain, K., Bao, M., Wang, Z.Z., Zhou, X., Guo, Z., Cao, M., et\u00a0al.: TheAgentCompany: Benchmarking LLM agents on consequential real world tasks. arXiv:2412.14161 (2024). (Sep 2025 update); Gemini\u00a02.5 Pro results (30.3%) from v3"},{"key":"9837_CR16","unstructured":"Tang, J., Fan, T., Huang, C.: AutoAgent: A fully-automated and zero-code framework for LLM agents. arXiv:2502.05957 (2025)"},{"key":"9837_CR17","unstructured":"Zhang, T., Pan, S., Zhang, Z., Xing, Z., Sun, X.: Deployability-centric infrastructure-as-code generation: An LLM-based iterative framework. arXiv:2506.05623 (2025)"},{"key":"9837_CR18","doi-asserted-by":"crossref","unstructured":"Fukuda, N., Nozue, H., Oishi, H.: Small language model agent for the operations of continuously updating ICT systems. IEEE Access (2025)","DOI":"10.1109\/ACCESS.2025.3544637"},{"key":"9837_CR19","doi-asserted-by":"crossref","unstructured":"Qian, C., Han, C., Fung, Y.R., Qin, Y., Liu, Z., Ji, H.: CREATOR: Tool creation for disentangling abstract and concrete reasoning of large language models. arXiv:2305.14318 (2023)","DOI":"10.18653\/v1\/2023.findings-emnlp.462"},{"key":"9837_CR20","doi-asserted-by":"crossref","unstructured":"W\u00f6lflein, G., Ferber, D., Truhn, D., Arandjelovi\u0107, O., Kather, J.N.: LLM agents making agent tools. arXiv:2502.11705 (2025)","DOI":"10.18653\/v1\/2025.acl-long.1266"},{"key":"9837_CR21","doi-asserted-by":"crossref","unstructured":"Takerngsaksiri, W., Pasuksmit, J., Thongtanunam, P., Tantithamthavorn, C., Zhang, R., Jiang, F., Li, J., Cook, E., Chen, K., Wu, M.: Human-in-the-loop software development agents. arXiv:2411.12924 (2025)","DOI":"10.1109\/ICSE-SEIP66354.2025.00036"},{"key":"9837_CR22","unstructured":"Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., Cao, Y.: ReAct: Synergizing reasoning and acting in language models. In: International Conference on Learning Representations (2023)"},{"key":"9837_CR23","unstructured":"LangChain. Langchain, (2025). Accessed: 16 Apr 2026"},{"key":"9837_CR24","doi-asserted-by":"crossref","unstructured":"Lin, X.V., Wang, C., Zettlemoyer, L., Ernst, M.D.: NL2Bash: A corpus and semantic parser for natural language interface to the linux operating system. arXiv:1802.08979 (2018)","DOI":"10.63317\/42er6x4bacdf"},{"key":"9837_CR25","unstructured":"LangGraph.: Langgraph (2026). Accessed: 16 Apr 2026"},{"key":"9837_CR26","unstructured":"LibreChat.: LibreChat: Enhanced ChatGPT clone (2024). https:\/\/github.com\/danny-avila\/LibreChat. Accessed: 16 Apr 2026"},{"key":"9837_CR27","unstructured":"Seyedkazemi\u00a0Ardebili, M.: kube_q: A terminal-based CLI client for KubeIntellect (2025). https:\/\/github.com\/MSKazemi\/kube_q. Accessed: 16 Apr 2026"},{"key":"9837_CR28","unstructured":"Lin, C., Han, Z., Zhang, C., Yang, Y., Yang, F., Chen, C., Qiu, L.: Parrot: Efficient serving of LLM-based applications with semantic variable. arXiv:2405.19888 (2024)"},{"key":"9837_CR29","doi-asserted-by":"crossref","unstructured":"Yang, J., Jimenez, C.E., Wettig, A., Lieret, K., Yao, S., Narasimhan, K., Press, O.: SWE-agent: Agent-computer interfaces enable automated software engineering. In: Advances in Neural Information Processing Systems (NeurIPS) (2024)","DOI":"10.52202\/079017-1601"},{"key":"9837_CR30","unstructured":"Kubernetes.: Kubernetes Python client (2025). Accessed: 16 Apr 2026"},{"key":"9837_CR31","unstructured":"Langfuse.: Langfuse: Open source LLM engineering platform (2024). https:\/\/github.com\/langfuse\/langfuse. Accessed: 16 Apr 2026"},{"key":"9837_CR32","unstructured":"Prometheus Authors.: Prometheus: Monitoring system and time series database (2012). https:\/\/prometheus.io. Accessed: 16 Apr 2026"},{"key":"9837_CR33","unstructured":"Ram\u00edrez, S.: FastAPI: Modern, fast web framework for building APIs (2019). https:\/\/fastapi.tiangolo.com. Accessed: 16 Apr 2026"},{"key":"9837_CR34","unstructured":"Jimenez, C.E., Yang, J., Wettig, A., Yao, S., Pei, K., Press, O., Narasimhan, K.: SWE-bench: Can language models resolve real-world GitHub issues? In: International Conference on Learning Representations (ICLR) (2024)"},{"key":"9837_CR35","unstructured":"Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Letman, A., et\u00a0al.: The Llama 3 herd of models. arXiv:2407.21783 (2024)"},{"key":"9837_CR36","unstructured":"Google Cloud.: GPU pricing (2026). https:\/\/cloud.google.com\/compute\/gpus-pricing. Accessed: 31 May 2026"},{"key":"9837_CR37","unstructured":"Microsoft.: Azure OpenAI in microsoft foundry models: Quotas and limits (2026). https:\/\/learn.microsoft.com\/en-us\/azure\/ai-services\/openai\/quotas-limits. Accessed: 16 Apr 2026"}],"container-title":["Journal of Grid Computing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10723-026-09837-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10723-026-09837-6","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10723-026-09837-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,27]],"date-time":"2026-06-27T10:43:18Z","timestamp":1782556998000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10723-026-09837-6"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,27]]},"references-count":37,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2026,9]]}},"alternative-id":["9837"],"URL":"https:\/\/doi.org\/10.1007\/s10723-026-09837-6","relation":{},"ISSN":["1570-7873","1572-9184"],"issn-type":[{"value":"1570-7873","type":"print"},{"value":"1572-9184","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,27]]},"assertion":[{"value":"21 August 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 June 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"27 June 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"17"}}