Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

ARCeR: an Agentic RAG for the Automated Definition of Cyber Ranges

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read ARCeR, an agentic retrieval-augmented generation system, turns natural-language requests into validated, deployable cyber range configurations, beating a plain LLM and a basic RAG system in controlled tests.

desk verdict Useful agentic RAG engineering for cyber range config generation; framework-agnostic claim overreaches and evidence is thin, but it deserves peer review. read the letter →

arxiv 2504.12143 v1 pith:TKI54EJQ submitted 2025-04-16 cs.CR cs.AI

Matteo Lupinacci , Francesco Blefari , Francesco Romeo , Francesco Aurelio Pironti , Angelo Furfaro This is my paper · ORCID
classification cs.CRcs.AI
keywords cyberrangesagenticRAGretrieval-augmentedgenerationLLMagentsnaturallanguagetoconfigurationcybersecuritytrainingsyntaxverificationself-correctionloop
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes ARCeR, an agentic retrieval-augmented generation system that takes a natural-language description of a training scenario and produces a configuration file a cyber range platform can validate and deploy. The authors' claim is that adding a retrieval step, a runtime syntax checker, and a self-correction loop turns a large language model into a dependable generator of cyber range definitions, where a plain LLM fails completely and a basic RAG system still produces invalid files. In controlled tests ARCeR handled 10 of 10 simple prompts and 18 of 20 more complex scenarios at the syntactic level, with 13 of 20 fully matching the request after manual semantic review. The authors further claim the design is framework-agnostic: retargeting to another cyber range platform only requires swapping the documents in the knowledge base. The payoff, if true, is that instructors could describe a range in plain language and get a working, deployable environment without hand-writing platform-specific configuration files.

What carries the argument

Agentic RAG loop. ARCeR couples a large-language-model reasoning core with two tools: a RAG subsystem that stores per-framework documentation chunks in separate vector stores and retrieves them with Maximal Marginal Relevance, taking 20 candidate chunks and filtering to the 8 most diverse and relevant, and a Checker Tool that submits the generated configuration to the target framework's syntax-validation interface and returns any error messages. The agent writes a draft, the checker validates it, and the LLM rewrites the draft in light of the reported errors, repeating the cycle up to three times. The same loop, together with the agent's memory across turns, is what lets ARCeR fix syntax errors automatically, resolve missing mandatory parameters by asking the user, and generate self-devised scenarios.

What would settle it

Run ARCeR against a second cyber range framework whose documentation is sparse or incomplete and which has no syntax-check API, using the same 20 natural-language scenarios; if valid, deployable outputs fall materially below the 90% syntactic and 65% fully correct rates reported for the tested framework, the framework-agnostic claim is refuted.

Watch

Extended reading notes

Core claim

The paper's core discovery is that an agentic loop—generate a draft configuration, run it through the framework's own syntax checker, feed error messages back to the language model, and retry—is what closes the gap between unreliable generation and usable output. On the CyRIS platform, the base LLM produced zero valid configurations on ten simple prompts even when given syntax details; a pure RAG system produced six; ARCeR produced ten. On twenty progressively harder scenarios covering all CyRIS features, ARCeR produced syntactically valid files for eighteen, and domain experts judged thirteen fully semantically correct, with the remaining five having only low- or medium-severity mismatches fixable by a follow-up user instruction. The paper also reports that ARCeR can autonomously invent a sensible network and host configuration from a bare prompt asking for a range on a given platform, which the authors position as an advance over prior random-scenario generators.

Load-bearing premise

The claim that ARCeR can target any cyber range framework rests on the assumption that each framework has both thorough documentation of its configuration syntax and a programmatic syntax checker that returns error messages, because only one such well-documented framework was tested.

Editorial extensions

If this is right

  • A natural-language description of a training range can be turned into a validated configuration file without hand-writing platform syntax.
  • The self-correction loop turns syntax feedback from the deployment platform into immediate retries, so errors that would block a basic RAG system are repaired automatically.
  • Adding support for a new cyber range framework becomes a documentation task rather than a model retraining task: prepare reference documents, load them into a separate vector store, and expose a syntax-checking endpoint.
  • Users can iteratively refine a generated range through conversation, because the agent keeps memory of previous requests, and missing mandatory parameters can be resolved by asking the user rather than failing.
  • Given only a platform name, the system can propose its own coherent training scenario, complete with machines, software, and network topology.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the framework-agnostic claim holds, the same generate-check-retry pattern could transfer to other declarative configuration tasks—cloud orchestration, container stacks, network policies, or CI pipelines—wherever an executable validator exists.
  • The case study required a manual curation pass reducing the framework documentation to 28 pages, so the hidden work behind 'any framework' may be producing reference documents of comparable density and correctness.
  • An ablation study separating retrieval quality from the checker retry loop would identify which component produces the jump from 6 to 10 on simple prompts.
  • A direct extension would test whether semantic errors can also be caught automatically by feeding deployment-time or post-boot checks back into the same agentic loop, not just syntactic validation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents ARCeR, an agentic retrieval-augmented generation (RAG) system that takes natural-language descriptions of a desired cyber range and produces configuration files for a target cyber range framework, with optional automated deployment. The system is built on an LLM reasoning engine, a RAG subsystem that retrieves chunks from framework-specific documents using maximal marginal relevance, and a checker tool that invokes an external syntax-checking API and feeds error messages back into the agent for iterative self-correction. The authors evaluate ARCeR using the CyRIS framework. A comparative test over 10 simple scenarios reports that a base LLM succeeded 0/10 times, a pure RAG system 6/10, and ARCeR 10/10. A second test with 20 progressively more complex scenario descriptions reports 18/20 syntactic successes (90%) and 13/20 fully correct outputs (65%), with a manual error-severity analysis of the remaining cases. The paper claims that ARCeR is framework-agnostic, in the sense that adapting to a different cyber range platform only requires swapping the documents in the vector store.

Significance. If the claims were fully substantiated, ARCeR would be a useful contribution to cyber range automation: it would lower the expertise barrier for instructors who need to translate training objectives into platform-specific configuration files, and it would demonstrate a concrete application of agentic RAG in a cybersecurity setting. The paper has real strengths: the comparative design (base LLM vs. RAG vs. agentic RAG) targets the right question, the use of a real framework with a syntax checker makes the self-correction loop concrete, the error-severity taxonomy (high/medium/low) is a sensible evaluation instrument, and Section 7 candidly states a significant limitation. However, the evidence base is thin: 30 prompts in total, no released prompts or code, no statistical tests, semantic evaluation performed manually by the authors without blinding or inter-rater reliability, and only one target framework tested. The central 'any CR framework' claim rests on an untested precondition that the target platform exposes a syntax-checking API. The contribution is therefore promising but currently overclaimed; the paper needs either substantially more evidence or a more modest framing.

major comments (4)
  1. [Abstract; Section 3.1; Section 4] The claim that ARCeR can target 'any CR framework' by simply changing the external documents is not supported by the evaluation. The self-correction loop in Section 3.1 (steps 2 and 3) depends on an external server exposing a syntax-checking API for the target platform; without such an API, the agent cannot perceive errors and degrades to a pure RAG system, losing exactly the advantage measured in Table 1 (10/10 vs. 6/10). Only CyRIS is tested, and Section 4 selects CyRIS partly because it has detailed documentation and a syntax checker. The Section 7 limitation that ARCeR cannot determine a priori whether a request exceeds framework capabilities further shows that framework adaptation is not reducible to document substitution. The framework-agnostic claim is an extrapolation from one favorable instance and should be either removed or supported by a second framework with an explicit statement of the checker-API precondition.
  2. [Section 5.3; Figure 3; Table 3] The central empirical claim rests on 20 manually written prompts with a 65% fully correct rate and 90% syntactic success, but semantic correctness was assessed by the authors themselves, with no blinding, no inter-rater reliability measure, and no statistical analysis. With 20 trials, the difference between 13/20 and 18/20 is not statistically meaningful without further reporting; the binomial confidence intervals overlap considerably. The paper should release the 20 prompt descriptions and the corresponding outputs, report per-scenario results, and ideally include an independent or blinded evaluation of semantic correctness. Without this, the quantitative claims in the abstract and conclusion are difficult to verify.
  3. [Section 5.2; Table 1] The comparison between the pure RAG system and ARCeR is not apples-to-apples. ARCeR's 10/10 result includes one case in which the agent used human-in-the-loop interaction to resolve missing mandatory parameters, while the pure RAG configuration was not given an equivalent interaction channel. In addition, the paper does not specify what 'additional details regarding the CyRIS framework syntax' were provided to the base LLM in the 0/10 condition, nor how the prompts were constructed. These details are necessary to interpret the comparison and to rule out that the base-LLM failure is due to an unfair or underspecified prompt rather than to a genuine lack of framework knowledge.
  4. [Section 7] The admitted limitation that ARCeR cannot determine a priori whether the user's request is within the capabilities of the target framework is load-bearing for the generality claim. For example, the system cannot detect that CyRIS only implements a bus topology and will attempt to generate a configuration that later fails during instantiation. This shows that a correct framework adaptation requires a model of framework capabilities, not just document retrieval. The paper should either present a mechanism for capability-aware validation or explicitly narrow the claimed scope to frameworks whose feature set is fully documented and whose syntax checker is available.
minor comments (6)
  1. [Abstract; Section 7] Minor typos: 'state-of-art' should be 'state-of-the-art' in the abstract, and 'a Agentic RAG' should be 'an Agentic RAG' in the conclusion.
  2. [Section 5.3; Figure 3] Figure 3 is difficult to parse: the left pie chart labels 'Semantics errors 25.0%, Failed 10.0%, Correct 65.0%', while the right chart reports 'Medium 40.0% Low 60.0%', but the right chart is not clearly defined as the distribution of the five semantic-error cases. Please label both charts fully and explain the relationship between the two percentages.
  3. [Section 5.1] There is a duplicated phrase: 'the way external tools are invoked both are both specific to the employed LLM' should read 'the way external tools are invoked are both specific to the employed LLM'.
  4. [Section 4] The manual curation of the CyRIS documentation is itself part of the adaptation effort: the authors removed installation and performance sections and retained 28 pages. This should be acknowledged as a human cost in the framework-adaptation discussion, since 'swapping documents' is not fully automatic.
  5. [Section 3.2; Section 5.3] The authors state that the tool will be made available open-source and that the tests use basic subscription accounts, but no repository link, prompt set, or artifact is provided. Releasing the prompts, generated files, and evaluation code would substantially strengthen reproducibility.
  6. [Section 5.2] In Table 1, the row 'Failure reasons' lists multiple items without indicating how many of the 10 failures correspond to each reason for the base LLM condition. A per-condition breakdown of failure types would be more informative.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity; the central claims are empirical comparisons, and the framework-generality claim is a conditional extrapolation rather than a circular derivation.

full rationale

ARCeR's central contribution is an implemented system and an empirical evaluation, not a mathematical derivation. The reported success rates (Section 5.2: 10/10 vs 6/10 vs 0/10; Section 5.3: 18/20 syntactic, 13/20 semantic) are measured against the external CyRIS syntax checker and manual expert review. No parameter is fitted to the benchmark outputs, and the checker is not an output of the model: it is an external tool whose error messages the agent must correctly interpret in order to repair the file. Failures still occur (2/20), so passing the checker is not guaranteed by construction. The only self-citations (refs [9,10]) appear in Section 2.1 as examples of real-time analysis frameworks for cyber ranges and are not load-bearing for any claim about ARCeR. The statement that ARCeR can target any CR framework 'provided that specific knowledge is made available to it' is an untested generalization with a hidden precondition (a programmatic syntax-checker API), but this is an overclaim or correctness risk, not circularity: the paper does not define 'target any framework' in terms of the evaluation outcome, nor does it derive the generalization from the CyRIS results. The comparison between pure RAG and ARCeR intentionally differs by the presence of the checker and user interaction; that is a design choice, not a self-referential derivation.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

No new physical or mathematical entities are introduced; the software components (checker tool, vector store, agent loop) are standard in agentic RAG systems. The free parameters are hand-chosen retrieval and evaluation settings, not fitted constants.

free parameters (2)
  • RAG retrieval hyperparameters = lambda_mult=0.5, chunk_size=1000, overlap=200, 20 chunks retrieved, 8 selected
    These values are chosen by the authors based on prior MMR practice and manual tuning; they affect retrieval diversity and thus the generated configs, but the paper does not justify them systematically or explore sensitivity.
  • Maximum retry attempts = 3
    The evaluation counts a test as successful if completed within three agent iterations; this threshold is an arbitrary evaluation choice that caps the observed success rate.
assumptions (3)
  • domain assumption CyRIS documentation and example files contain sufficient information for the LLM to generate valid configuration files for any requested scenario within CyRIS capabilities.
    Section 4 states the RAG knowledge base is built from the CyRIS paper, user guide, and six description files after manual cleanup; the system's success depends on this documentation being complete and correctly chunked.
  • domain assumption The syntax checker exposed by the CR platform returns error messages complete and accurate enough for the agent to correct its output.
    Section 3.1 describes the checker tool feedback loop; if the checker is partial or its messages are not diagnostic, self-correction would not converge as claimed.
  • domain assumption Natural language descriptions by users can be mapped by the LLM to the CyRIS feature set without ambiguity.
    The evaluation uses scenarios written by domain experts; the paper notes unresolved cases where mandatory parameters were missing, requiring user clarification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ARCeR: an Agentic RAG for the Automated Definition of Cyber Ranges." pith.science (2026). https://pith.science/paper/TKI54EJQ

@misc{pith2026250412143,
  author       = {Pith},
  title        = {Pith review of: ARCeR: an Agentic RAG for the Automated Definition of Cyber Ranges},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TKI54EJQ}},
  note         = {Machine review of arXiv:2504.12143}
}
read the original abstract

The growing and evolving landscape of cybersecurity threats necessitates the development of supporting tools and platforms that allow for the creation of realistic IT environments operating within virtual, controlled settings as Cyber Ranges (CRs). CRs can be exploited for analyzing vulnerabilities and experimenting with the effectiveness of devised countermeasures, as well as serving as training environments for building cyber security skills and abilities for IT operators. This paper proposes ARCeR as an innovative solution for the automatic generation and deployment of CRs, starting from user-provided descriptions in a natural language. ARCeR relies on the Agentic RAG paradigm, which allows it to fully exploit state-of-art AI technologies. Experimental results show that ARCeR is able to successfully process prompts even in cases that LLMs or basic RAG systems are not able to cope with. Furthermore, ARCeR is able to target any CR framework provided that specific knowledge is made available to it.

Figures

Figures reproduced from arXiv: 2504.12143 by the authors.

Figure 1
Figure 1. AI agent structure [16] end, the LLM can decide whether the generated answer is sufficient or if more work is needed. 2.3 Agentic RAG Despite their flexibility, general purpose LLMs, and the related LLM agents, often lack the domain-specific knowledge required to solve particular and non￾trivial tasks. Such a problem could be solved by re-training or fine-tuning the model; however, the cost of these operations is no… view at source ↗
Figure 2
Figure 2. Overall approach schema Given ARCeR’s support to interact with disparate CR frameworks it em￾ploys different vector stores to separately save the embeddings’ chunks doc￾uments related to different platforms. All documents (e.g., framework usage guides and example configuration files) are stored in a database containing a folder whose name coincides with the name of the specific CR framework. The RAG subsystem exploi… view at source ↗
Figure 3
Figure 3. Performance analysis The results obtained are presented in [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. ICSLure: A Very High Interaction Honeynet for PLC-based Industrial Control Systems

    cs.CR 2025-09 conditional novelty 6.0 of 10

    A hybrid honeynet that pairs real PLC hardware with a real-time physics simulator to lure and log ICS attackers.

Reference graph

Works this paper leans on

36 extracted references · 28 canonical work pages · cited by 1 Pith paper

  1. [1]

    Toward next- generation cyber range: A comparative study of training platforms,

    A. Grimaldi, J. Ribiollet, P. Nespoli, and J. Garcia-Alfaro, “Toward next- generation cyber range: A comparative study of training platforms,” in Lecture Notes in Computer Science . Springer Nature Switzerland, 2024, pp. 271–290. [Online]. Available: http://dx.doi.org/10.1007/978-3-031-54129-2_16

    work page doi:10.1007/978-3-031-54129-2_16 2024

  2. [2]

    Cyber range automation overview with a case study of CRATE,

    T. Gustafsson and J. Almroth, “Cyber range automation overview with a case study of CRATE,” inLecture Notes in Computer Science . Springer International Publishing, 2021, pp. 192–209. [Online]. Available: http://dx.doi.org/10.1007/ 978-3-030-70852-8_12

    work page 2021

  3. [3]

    Automating the deployment of cyber range with openstack,

    S. Zhou, J. He, T. Li, X. Lan, Y. Wang, H. Zhao, and Y. Li, “Automating the deployment of cyber range with openstack,”The Computer Journal , vol. 67, pp. 851–863, 2023. [Online]. Available: http://dx.doi.org/10.1093/comjnl/bxad024

    work page doi:10.1093/comjnl/bxad024 2023

  4. [4]

    Applications of llms for generating cyber security exercise scenarios,

    M. Mudassar Yamin, E. Hashmi, M. Ullah, and B. Katt, “Applications of llms for generating cyber security exercise scenarios,” IEEE Access, vol. 12, pp. 143806– 143822, 2024

    work page 2024

  5. [5]

    Agentic retrieval- augmented generation: A survey on agentic rag,

    A. Singh, A. Ehtesham, S. Kumar, and T. T. Khoei, “Agentic retrieval- augmented generation: A survey on agentic rag,” 2025. [Online]. Available: https://dx.doi.org/10.48550/ARXIV.2501.09136

    arXiv doi:10.48550/arxiv.2501.09136 2025

  6. [6]

    Cyexec*: A high-performance container-based cyber rangewithscenariorandomization,

    R. Nakata and A. Otsuka, “Cyexec*: A high-performance container-based cyber rangewithscenariorandomization,” IEEE Access,vol.9,pp.109095–109114,2021

    work page 2021

  7. [7]

    Cyber- security education and training support system: Cyris,

    R. Beuran, C. Pham, D. Tang, K.-i. Chinen, Y. Tan, and Y. Shinoda, “Cyber- security education and training support system: Cyris,” IEICE Transactions on Information and Systems , vol. E101.D, no. 3, pp. 740–749, 2018

    work page 2018

  8. [8]

    Cyber ranges,

    NIST, “Cyber ranges,” 2018. [Online]. Available: https://www.nist.gov/system/ files/documents/2018/02/13/cyber_ranges.pdf

    work page 2018

Show all 36 references
  1. [9]

    Toward a log-based anomaly detection system for cyber range platforms,

    F. Blefari, F. A. Pironti, and A. Furfaro, “Toward a log-based anomaly detection system for cyber range platforms,” in Proceedings of the 19th International Conference on Availability, Reliability and Security , ser. ARES ’24. New York, NY, USA: Association for Computing Machi...

    2024

  2. [10]

    Unveiling attack patterns from CTF network logs with process mining techniques,

    F. Romeo, F. Blefari, F. A. Pironti, and A. Furfaro, “Unveiling attack patterns from CTF network logs with process mining techniques,” in Proceedings of the Joint National Conference on Cybersecurity (ITASEC & SERICS 2025) , 2025

    2025

  3. [11]

    What is a cyber range? learn hands-on cybersecurity skills,

    H. Taylor, “What is a cyber range? learn hands-on cybersecurity skills,” https: //cybersecurityguide.org/resources/cyber-ranges/, 2023

    2023

  4. [12]

    Thecyberrange:Aguide,

    NIST,“Thecyberrange:Aguide,” 2023.[Online].Available:https://www.nist.gov/ system/files/documents/2023/09/29/The%20Cyber%20Range_A%20Guide.pdf

    2023

  5. [13]

    National cyber range overview,

    B. Ferguson, A. Tall, and D. Olsen, “National cyber range overview,” in2014 IEEE Military Communications Conference, 2014, pp. 123–128. 18 Lupinacci et al

    2014

  6. [14]

    Futureinternetresearchand experimentation,

    A.Gavras,A.Karila,S.Fdida,M.May,andM.Potts,“Futureinternetresearchand experimentation,” ACM SIGCOMM Computer Communication Review , vol. 37, pp. 89–92, 2007. [Online]. Available: http://dx.doi.org/10.1145/1273445.1273460

    2007

  7. [15]

    Wooldridge, An Introduction to MultiAgent Systems , 2nd ed

    M. Wooldridge, An Introduction to MultiAgent Systems , 2nd ed. Wiley, 2009

    2009

  8. [16]

    The rise and potential of large language model based agents: a survey,

    Z. Xi, W. Chen, X. Guo, W. He, Y. Ding, B. Hong, M. Zhang, J. Wang, S. Jin, E. Zhou, R. Zheng, X. Fan, X. Wang, L. Xiong, Y. Zhou, W. Wang, C. Jiang, Y. Zou, X. Liu, Z. Yin, S. Dou, R. Weng, W. Qin, Y. Zheng, X. Qiu, X. Huang, Q. Zhang, and T. Gui, “The rise and potential of l...

    2025

  9. [17]

    Retrieval-augmented generation for knowledge-intensive nlp tasks,

    P. Lewis, E. Perez, A. Piktus, F. Petroni, V. Karpukhin, N. Goyal, H. Küttler, M. Lewis, W.-t. Yih, T. Rocktäschelet al., “Retrieval-augmented generation for knowledge-intensive nlp tasks,”Advances in neural information processing systems, vol. 33, pp. 9459–9474, 2020

    2020

  10. [18]

    Langchain,

    H. Chase, “Langchain,” October 2022. [Online]. Available: https://github.com/ langchain-ai/langchain

    2022

  11. [19]

    Llamaindex,

    J. Liu, “Llamaindex,” November 2022. [Online]. Available: https://github.com/ jerryjliu/llama_index

    2022

  12. [20]

    Langdroid

    P. Chalasani and S. Jha, “Langdroid.” [Online]. Available: https://github.com/ langroid/langroid

  13. [21]

    C. E. Ingy döt Net and O. Ben-Kiki., “Yaml,” 2001. [Online]. Available: https://yaml.org/about.html

    2001

  14. [22]

    Summarization: (1) using MMR for diversity- based reranking and (2) evaluating summaries,

    J. Goldstein and J. Carbonell, “Summarization: (1) using MMR for diversity- based reranking and (2) evaluating summaries,” inTIPSTER TEXT PROGRAM PHASE III: Proceedings of a Workshop held at Baltimore, Maryland, October 13-15, 1998 . Association for Computational Linguistics, ...

    1998

  15. [23]

    Claude 3.7 sonnet system card,

    Anthropic, “Claude 3.7 sonnet system card,” 2025. [On- line]. Available: https://assets.anthropic.com/m/785e231869ea8b3b/original/ claude-3-7-sonnet-system-card.pdf

    2025

  16. [24]

    Sentence-BERT: Sentence embeddings using Siamese BERT-networks,

    N. Reimers and I. Gurevych, “Sentence-BERT: Sentence embeddings using Siamese BERT-networks,” in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, 11 2019, pp. 3982–3992

    2019

  17. [25]

    Is cosine-similarity of embeddings really about similarity?

    H. Steck, C. Ekanadham, and N. Kallus, “Is cosine-similarity of embeddings really about similarity?” inProc. of WWW ’24: The ACM Web Conference 2024 . ACM, 2024, pp. 887–890

    2024

  18. [26]

    LangGraph

    C. Nuno, B. Vadym, and F. William, “LangGraph.” [Online]. Available: https://github.com/langchain-ai/langgraph

  19. [27]

    cyb3rlab/cyris

    R. Beuran, “cyb3rlab/cyris.” [Online]. Available: https://github.com/cyb3rlab/ cyris

  20. [28]

    Gpt-4o system card,

    O. 2024, “Gpt-4o system card,” 2024. [Online]. Available: https://arxiv.org/abs/ 2410.21276

    2024 arXiv

  21. [29]

    Mistral large 2,

    M. A. team, “Mistral large 2,” 2024. [Online]. Available: https://mistral.ai/news/ mistral-large-2407

    2024

  22. [30]

    Nautilus: A tool for automated de- ployment and sharing of cyber range scenarios,

    G. Bernardinetti, S. Iafrate, and G. Bianchi, “Nautilus: A tool for automated de- ployment and sharing of cyber range scenarios,”Proceedings of the 16th Interna- tional Conference on Availability, Reliability and Security , pp. 1–7, 2021

    2021

  23. [31]

    Automating the generation of cyber range virtual scenarios with vsdl,

    G. Costa, E. Russo, and A. Armando, “Automating the generation of cyber range virtual scenarios with vsdl,”arXiv preprint arXiv:2001.06681 , 2020. ARCeR: an Agentic RAG for Cyber Ranges 19

    2001 arXiv

  24. [32]

    When LLMs meet cybersecurity: a systematic literature review,

    J.Zhang,H.Bu,H.Wen,Y.Liu,H.Fei,R.Xi,L.Li,Y.Yang,H.Zhu,andD.Meng, “When LLMs meet cybersecurity: a systematic literature review,”Cybersecurity,

  25. [33]

    LLM agents can autonomously hack websites,

    R. Fang, R. Bindu, A. Gupta, Q. Zhan, and D. Kang, “LLM agents can autonomously hack websites,” arXiv, 2024. [Online]. Available: https: //arxiv.org/abs/2402.06664

    2024 arXiv

  26. [34]

    LLM agents can autonomously exploit one-day vulnerabilities,

    R. Fang, R. Bindu, A. Gupta, and D. Kang, “LLM agents can autonomously exploit one-day vulnerabilities,” arXiv preprint arXiv:2404.08144 , 2024

    2024 arXiv

  27. [35]

    Depending on yourself when you should: Men- toring LLM with RL agents to become the master in cybersecurity games,

    Y. Yan, Y. Zhang, and K. Huang, “Depending on yourself when you should: Men- toring LLM with RL agents to become the master in cybersecurity games,”arXiv preprint arXiv:2403.17674, 2024

    2024 arXiv

  28. [2025]

    Available: http://dx.doi.org/10.1186/s42400-025-00361-w

    [Online]. Available: http://dx.doi.org/10.1186/s42400-025-00361-w

    doi

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.