---
title: "Why Do Multi-Agent LLM Systems Fail? — Cemri 2025"
description: "Multi-agent LLM systems (MAS) are popular, yet they often perform little better than single agents or… 84 references, cited by 71 papers here."
authors: ["Mert Cemri", "Melissa Z. Pan", "Shuyi Yang", "Lakshya A. Agrawal", "Bhavya Chopra", "Rishabh Tiwari", "Kurt Keutzer", "Aditya Parameswaran", "Dan Klein", "Kannan Ramchandran", "Matei Zaharia", "Joseph E. Gonzalez", "Ion Stoica"]
year: 2025
venue: "arXiv:2503.13657"
source: "https://arxiv.org/abs/2503.13657"
doi: "10.48550/arxiv.2503.13657"
arxiv: "2503.13657"
licence: "CC BY-NC-ND 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by-nc-nd/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/cemri-2025-why-multi-agent-llm/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/cemri-2025-why-multi-agent-llm/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/) / Why Do Multi-Agent LLM Systems Fail?

# Why Do Multi-Agent LLM Systems Fail?

**Authors:** Mert Cemri, Melissa Z. Pan, Shuyi Yang, Lakshya A. Agrawal, Bhavya Chopra, Rishabh Tiwari, [Kurt Keutzer](https://md.research.elamur.ai/free/en/authors/kurt-keutzer/), Aditya Parameswaran, [Dan Klein](https://md.research.elamur.ai/free/en/authors/dan-klein/), Kannan Ramchandran, [Matei Zaharia](https://md.research.elamur.ai/free/en/authors/matei-zaharia/), [Joseph E. Gonzalez](https://md.research.elamur.ai/free/en/authors/joseph-e-gonzalez/) and [Ion Stoica](https://md.research.elamur.ai/free/en/authors/ion-gabriel-stoica/) · **Year:** 2025 · **Venue:** arXiv:2503.13657 · **Licence:** [CC BY-NC-ND 4.0 (arXiv)](http://creativecommons.org/licenses/by-nc-nd/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2503.13657) · [PDF](https://arxiv.org/pdf/2503.13657) · [DOI 10.48550/arxiv.2503.13657](https://doi.org/10.48550/arxiv.2503.13657) · [arXiv 2503.13657](https://arxiv.org/abs/2503.13657) · [OpenAlex](https://openalex.org/W6910824972)

**Citations:** 18 citations (OpenAlex, 2026-10-14) · [71](https://md.research.elamur.ai/free/en/papers/cemri-2025-why-multi-agent-llm/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Multi-agent LLM systems (MAS) are popular, yet they often perform little better than single agents or simple baselines. The authors measured failure rates of 41% to 86.7% across seven open-source frameworks. They set out to explain these failures systematically. Six experts analyzed more than 150 long execution traces from five frameworks using Grounded Theory. From this they built MAST, a taxonomy of 14 failure modes in three groups: system design issues, inter-agent misalignment and task verification. To label data at scale, they calibrated an LLM-as-a-Judge annotator based on OpenAI's o1 against human labels. They then used it to produce MAST-Data, a released corpus of 1642 annotated traces covering coding, math and general-agent tasks. Failure profiles differ by framework and model. Simple workflow or prompt fixes help only partly, which suggests that many failures come from how agents are organized rather than from weak base models. The dataset, taxonomy and annotator give developers a shared vocabulary for diagnosing problems and comparing MAS designs.

## Key points
- **MAST-Data** contains 1642 traces from ChatDev, MetaGPT, HyperAgent, AppWorld, AG2 (MathChat), Magentic-One and OpenManus. The models include GPT-4o, Claude-3.7-Sonnet, Qwen2.5-Coder-32B-Instruct and CodeLlama-7b-Instruct-hf. A separate set, **MAST-Data-human**, has 21 traces labeled by three experts each.
- Human annotators reached κ = 0.88 on the taxonomy definitions. The o1-based annotator agreed with humans at κ = 0.77 and scored κ = 0.79 on two unseen systems and benchmarks.
- System design issues were the largest category at 44.2%, followed by inter-agent misalignment (32.3%) and task verification (23.5%). The most frequent single modes were step repetition (15.7%), reasoning-action mismatch (13.2%) and disobeying the task specification (11.8%).
- In ChatDev, giving the CEO agent the final decision raised task success by 9.4%. Across the intervention case studies, the best gain with the same model was 15.6%, but completion rates stayed low.
- The resources are available on GitHub, on Hugging Face (`mcemri/MAST-Data`) and as a Python package (`pip install agentdash`).

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works

## Cite

Cite as: Mert Cemri, Melissa Z. Pan, Shuyi Yang, Lakshya A. Agrawal, Bhavya Chopra, Rishabh Tiwari, Kurt Keutzer, Aditya Parameswaran, Dan Klein, Kannan Ramchandran, Matei Zaharia, Joseph E. Gonzalez and Ion Stoica. 2025. Why Do Multi-Agent LLM Systems Fail?. arXiv:2503.13657.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{cemri2025why,
  title = {{Why Do Multi-Agent LLM Systems Fail?}},
  author = {Mert Cemri, Melissa Z. Pan, Shuyi Yang, Lakshya A. Agrawal, Bhavya Chopra, Rishabh Tiwari, Kurt Keutzer, Aditya Parameswaran, Dan Klein, Kannan Ramchandran, Matei Zaharia, Joseph E. Gonzalez and Ion Stoica},
  year = {2025},
  doi = {10.48550/arxiv.2503.13657},
  eprint = {2503.13657},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2503.13657}
}
```

- [← previous SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines](https://md.research.elamur.ai/free/en/papers/team-2025-supergpqa-scaling-llm-evaluation-across/)
- [next → ReTool: Reinforcement Learning for Strategic Tool Use in LLMs](https://md.research.elamur.ai/free/en/papers/feng-2025-retool-reinforcement-learning-strategic-tool/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/)

## References

84 references, 32 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Leo Tolstoy. Anna Karenina. The Russian Messenger, 1878. [find](https://scholar.google.com/scholar?q=Leo+Tolstoy%2E+Anna+Karenina%2E+The+Russian+Messenger%2C+1878%2E)
2. Shishir G. Patil, Tianjun Zhang, Xin Wang, and Joseph E. Gonzalez. [Gorilla: Large language model connected with massive apis](https://md.research.elamur.ai/free/en/papers/patil-2023-gorilla-large-language-model-connected/), 2023. URL https://arxiv.org/abs/2305.15334. [↗](https://arxiv.org/abs/2305.15334)
3. Charles Packer, Sarah Wooders, Kevin Lin, Vivian Fang, Shishir G. Patil, Ion Stoica, and Joseph E. Gonzalez. Memgpt: Towards llms as operating systems, 2024. URL https://arxiv.org/abs/2310.08560. [↗](https://arxiv.org/abs/2310.08560)
4. Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, Wayne Xin Zhao, Zhewei Wei, and Jirong Wen. [A survey on large language model based autonomous agents](https://md.research.elamur.ai/free/en/papers/wang-2023-survey-large-language-model-based/). Frontiers of Computer Science, 18(6), March 2024. ISSN 2095-2236. doi: 10.1007/s…
5. Chen Qian, Wei Liu, Hongzhang Liu, Nuo Chen, Yufan Dang, Jiahao Li, Cheng Yang, Weize Chen, Yusheng Su, Xin Cong, Juyuan Xu, Dahai Li, Zhiyuan Liu, and Maosong Sun. [Chatdev: Communicative agents for software development](https://md.research.elamur.ai/free/en/papers/qian-2023-chatdev-communicative-agents-software/). arXiv preprint arXiv:2307.07924, 2023. URL https://arxiv.org/abs/2307.07924. [↗](https://arxiv.org/abs/2307.07924)
6. Xingyao Wang, Boxuan Li, Yufan Song, Frank F. Xu, Xiangru Tang, Mingchen Zhuge, Jiayi Pan, Yueqi Song, Bowen Li, Jaskirat Singh, Hoang H. Tran, Fuqiang Li, Ren Ma, Mingzhang Zheng, Bill Qian, Yanjun Shao, Niklas Muennighoff, Yizhe Zhang, Binyuan Hui, Junyang Lin, Robert Brennan, Hao Peng, Heng Ji, a… — [OpenHands: An Open Platform for AI Software Developers as Generalist Agents](https://md.research.elamur.ai/free/en/papers/wang-2024-openhands-open-platform-ai/)
7. Juraj Gottweis, Wei-Hung Weng, Alexander Daryin, Tao Tu, Anil Palepu, Petar Sirkovic, Artiom Myaskovsky, Felix Weissenberger, Keran Rong, Ryutaro Tanno, Khaled Saab, Dan Popovici, Jacob Blum, Fan Zhang, Katherine Chou, Avinatan Hassidim, Burak Gokturk, Amin Vahdat, Pushmeet Kohli, Yossi Matias, Andr… — [Accelerating scientific discovery with Co-Scientist](https://md.research.elamur.ai/free/en/papers/gottweis-2025-accelerating-scientific-discovery-co/)
8. Kyle Swanson, Wesley Wu, Nash L. Bulaong, John E. Pak, and James Zou. The virtual lab: Ai agents design new sars-cov-2 nanobodies with experimental validation. bioRxiv, 2024. doi: 10.1101/2024.11.11.623004. URL [https://www.biorxiv.org/content/early/2024/11/12/2024.11.11.623004](https://www.biorxiv.org/content/early/2024/11/12/2024.11.11.623004). [↗](https://www.biorxiv.org/content/early/2024/11/12/2024.11.11.623004)
9. Joon Sung Park, Joseph C. O’Brien, Carrie J. Cai, Meredith Ringel Morris, Percy Liang, and Michael S. Bernstein. Generative agents: Interactive simulacra of human behavior, 2023. URL https://arxiv.org/abs/2304.03442. [↗](https://arxiv.org/abs/2304.03442)
10. Xinbin Liang, Jinyu Xiang, Zhaoyang Yu, Jiayi Zhang, and Sirui Hong. Openmanus: An open-source framework for building general ai agents. [https://github.com/mannaandpoem/OpenManus](https://github.com/mannaandpoem/OpenManus), 2025. [↗](https://github.com/mannaandpoem/OpenManus)
11. Adam Fourney, Gagan Bansal, Hussein Mozannar, Cheng Tan, Eduardo Salinas, Friederike Niedtner, Grace Proebsting, Griffin Bassman, Jack Gerrits, Jacob Alber, et al. Magentic-one: A generalist multi-agent system for solving complex tasks. arXiv preprint arXiv:2411.04468, 2024. [find](https://scholar.google.com/scholar?q=Adam+Fourney%2C+Gagan+Bansal%2C+Hussein+Mozannar%2C+Cheng+Tan%2C+Eduardo+Salinas%2C+Friederike+Niedtner%2C+Grace+Proebsting%2C+Griffin+Bassman%2C+Jack+Gerrits%2C+Jacob+Alber%2C+et+al%2E+Magentic%2Done%3A+A+generalist+multi%2Dage)
12. Junda He, Christoph Treude, and David Lo. Llm-based multi-agent systems for software engineering: Vision and the road ahead, 2024. URL https://arxiv.org/abs/2404.04834. — [arXiv:2404.04834](https://arxiv.org/abs/2404.04834) [↗](https://arxiv.org/abs/2404.04834)
13. Zhao Mandi, Shreeya Jain, and Shuran Song. Roco: Dialectic multi-robot collaboration with large language models, 2023. URL https://arxiv.org/abs/2307.04738. — [arXiv:2307.04738](https://arxiv.org/abs/2307.04738) [↗](https://arxiv.org/abs/2307.04738)
14. Hongxin Zhang, Weihua Du, Jiaming Shan, Qinhong Zhou, Yilun Du, Joshua B. Tenenbaum, Tianmin Shu, and Chuang Gan. [Building cooperative embodied agents modularly with large language models](https://md.research.elamur.ai/free/en/papers/zhang-2023-building-cooperative-embodied-agents-modularly/), 2024. URL https://arxiv.org/abs/2307.02485. [↗](https://arxiv.org/abs/2307.02485)
15. Yilun Du, Shuang Li, Antonio Torralba, Joshua B. Tenenbaum, and Igor Mordatch. [Improving factuality and reasoning in language models through multiagent debate](https://md.research.elamur.ai/free/en/papers/du-2023-improving-factuality-reasoning-language-models/), 2023. URL https://arxiv.org/abs/2305.14325. [↗](https://arxiv.org/abs/2305.14325)
16. Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th annual acm symposium on user interface software and technology, pages 1–22, 2023. [find](https://scholar.google.com/scholar?q=Joon+Sung+Park%2C+Joseph+O%E2%80%99Brien%2C+Carrie+Jun+Cai%2C+Meredith+Ringel+Morris%2C+Percy+Liang%2C+and+Michael+S+Bernstein%2E+Generative+agents%3A+Interactive+simulacra+of+human+behavior%2E+In+Proceedings+of+the+36th+ann)
17. Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V Chawla, Olaf Wiest, and Xiangliang Zhang. [Large language model based multi-agents: A survey of progress and challenges](https://md.research.elamur.ai/free/en/papers/guo-2024-large-language-model-based/). arXiv preprint arXiv:2402.01680, 2024.
18. Chunqiu Steven Xia, Yinlin Deng, Soren Dunn, and Lingming Zhang. [Agentless: Demystifying llm-based software engineering agents](https://md.research.elamur.ai/free/en/papers/xia-2024-agentless-demystifying-llm-based/), 2024. URL https://arxiv.org/abs/2407.01489. [↗](https://arxiv.org/abs/2407.01489)
19. Sayash Kapoor, Benedikt Stroebl, Zachary S. Siegel, Nitya Nadgir, and Arvind Narayanan. [Ai agents that matter](https://md.research.elamur.ai/free/en/papers/kapoor-2024-ai-agents-that-matter/), 2024. URL https://arxiv.org/abs/2407.01502. [↗](https://arxiv.org/abs/2407.01502)
20. Barney G. Glaser and Anselm L. Strauss. The Discovery of Grounded Theory: Strategies for Qualitative Research. Aldine Publishing Company, 1967. [find](https://scholar.google.com/scholar?q=Barney+G%2E+Glaser+and+Anselm+L%2E+Strauss%2E+The+Discovery+of+Grounded+Theory%3A+Strategies+for+Qualitative+Research%2E+Aldine+Publishing+Company%2C+1967%2E)
21. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric P. Xing, Hao Zhang, Joseph E. Gonzalez, and Ion Stoica. [Judging llm-as-a-judge with mt-bench and chatbot arena](https://md.research.elamur.ai/free/en/papers/zheng-2023-judging-llm-judge-mt-bench/), 2023. URL https://arxiv.org/abs/2306.05685. [↗](https://arxiv.org/abs/2306.05685)
22. Zora Zhiruo Wang, Jiayuan Mao, Daniel Fried, and Graham Neubig. [Agent workflow memory](https://md.research.elamur.ai/free/en/papers/wang-2024-agent-workflow-memory/), 2024. URL https://arxiv.org/abs/2409.07429. [↗](https://arxiv.org/abs/2409.07429)
23. Omar Khattab, Arnav Singhvi, Paridhi Maheshwari, Zhiyuan Zhang, Keshav Santhanam, Sri Vardhamanan, Saiful Haq, Ashutosh Sharma, Thomas T. Joshi, Hanna Moazam, Heather Miller, Matei Zaharia, and Christopher Potts. [Dspy: Compiling declarative language model calls into self-improving pipelines](https://md.research.elamur.ai/free/en/papers/khattab-2023-dspy-compiling-declarative-language-model/), 2023. U…
24. Yiran Wu, Tianwei Yue, Shaokun Zhang, Chi Wang, and Qingyun Wu. [Stateflow: Enhancing llm task-solving through state-driven workflows](https://md.research.elamur.ai/free/en/papers/wu-2024-stateflow-enhancing-llm-task-solving/), 2024. URL https://arxiv.org/abs/2403.11322. [↗](https://arxiv.org/abs/2403.11322)
25. Shanshan Han, Qifan Zhang, Yuhang Yao, Weizhao Jin, Zhaozhuo Xu, and Chaoyang He. Llm multi-agent systems: Challenges and open problems, 2024. URL https://arxiv.org/abs/2402.03578. — [arXiv:2402.03578](https://arxiv.org/abs/2402.03578) [↗](https://arxiv.org/abs/2402.03578)
26. Lewis Hammond, Alan Chan, Jesse Clifton, Jason Hoelscher-Obermaier, Akbir Khan, Euan McLean, Chandler Smith, Wolfram Barfuss, Jakob Foerster, Tomáš Gavenčiak, The Anh Han, Edward Hughes, Vojtěch Kovařı́k, Jan Kulveit, Joel Z. Leibo, Caspar Oesterheld, Christian Schroeder de Witt, Nisarg Shah, M… — [arXiv:2502.14143](https://arxiv.org/abs/2502.14143) [find](https://scholar.google.com/scholar?q=Lewis+Hammond%2C+Alan+Chan%2C+Jesse+Clifton%2C+Jason+Hoelscher%2DObermaier%2C+Akbir+Khan%2C+Euan+McLean%2C+Chandler+Smith%2C+Wolfram+Barfuss%2C+Jakob+Foerster%2C+Toma%CC%81s%CC%8C+Gavenc%CC%8Ciak%2C+The+Anh+Han%2C+Edward+Hughes%2C+Vojte%CC%8Cch+K)
27. Carlos E Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik R Narasimhan. [SWE-bench: Can language models resolve real-world github issues?](https://md.research.elamur.ai/free/en/papers/jimenez-2024-swe-bench-can-language-models/) In The Twelfth International Conference on Learning Representations, 2024. URL [https://openreview.net/forum?id=VTF8yNQM66](https://openreview.net/forum?id=VTF8yNQM66). [↗](https://openreview.net/forum?id=VTF8yNQM66)
28. Ji-Lun Peng, Sijia Cheng, Egil Diau, Yung-Yu Shih, Po-Heng Chen, Yen-Ting Lin, and Yun-Nung Chen. A survey of useful llm evaluation. arXiv preprint arXiv:2406.00936, 2024. — [arXiv:2406.00936](https://arxiv.org/abs/2406.00936) [find](https://scholar.google.com/scholar?q=Ji%2DLun+Peng%2C+Sijia+Cheng%2C+Egil+Diau%2C+Yung%2DYu+Shih%2C+Po%2DHeng+Chen%2C+Yen%2DTing+Lin%2C+and+Yun%2DNung+Chen%2E+A+survey+of+useful+llm+evaluation%2E+arXiv+preprint+arXiv%3A2406%2E00936%2C+2024%2E+%E2%80%94+%5BarXiv%3A2406%2E00936%5D%28https%3A%2F)
29. Wei Wang, Dan Zhang, Tao Feng, Boyan Wang, and Jie Tang. Battleagentbench: A benchmark for evaluating cooperation and competition capabilities of language models in multi-agent systems. arXiv preprint arXiv:2408.15971, 2024. — [arXiv:2408.15971](https://arxiv.org/abs/2408.15971) [find](https://scholar.google.com/scholar?q=Wei+Wang%2C+Dan+Zhang%2C+Tao+Feng%2C+Boyan+Wang%2C+and+Jie+Tang%2E+Battleagentbench%3A+A+benchmark+for+evaluating+cooperation+and+competition+capabilities+of+language+models+in+multi%2Dagent+systems%2E+arXiv+preprint)
30. Timothée Anne, Noah Syrkis, Meriem Elhosni, Florian Turati, Franck Legendre, Alain Jaquier, and Sebastian Risi. Harnessing language for coordination: A framework and benchmark for llm-driven multi-agent control. arXiv preprint arXiv:2412.11761, 2024. — [arXiv:2412.11761](https://arxiv.org/abs/2412.11761) [find](https://scholar.google.com/scholar?q=Timothe%CC%81e+Anne%2C+Noah+Syrkis%2C+Meriem+Elhosni%2C+Florian+Turati%2C+Franck+Legendre%2C+Alain+Jaquier%2C+and+Sebastian+Risi%2E+Harnessing+language+for+coordination%3A+A+framework+and+benchmark+for+llm%2Ddriven+multi%2Dag)
31. Matteo Bettini, Amanda Prorok, and Vincent Moens. Benchmarl: Benchmarking multi-agent reinforcement learning. Journal of Machine Learning Research, 25(217):1–10, 2024. [find](https://scholar.google.com/scholar?q=Matteo+Bettini%2C+Amanda+Prorok%2C+and+Vincent+Moens%2E+Benchmarl%3A+Benchmarking+multi%2Dagent+reinforcement+learning%2E+Journal+of+Machine+Learning+Research%2C+25%28217%29%3A1%E2%80%9310%2C+2024%2E)
32. Qian Long, Zhi Li, Ran Gong, Ying Nian Wu, Demetri Terzopoulos, and Xiaofeng Gao. Teamcraft: A benchmark for multi-modal multi-agent systems in minecraft. arXiv preprint arXiv:2412.05255, 2024. — [arXiv:2412.05255](https://arxiv.org/abs/2412.05255) [find](https://scholar.google.com/scholar?q=Qian+Long%2C+Zhi+Li%2C+Ran+Gong%2C+Ying+Nian+Wu%2C+Demetri+Terzopoulos%2C+and+Xiaofeng+Gao%2E+Teamcraft%3A+A+benchmark+for+multi%2Dmodal+multi%2Dagent+systems+in+minecraft%2E+arXiv+preprint+arXiv%3A2412%2E05255%2C+2024%2E+%E2%80%94+%5BarX)
33. Yang Liu, Yuanshun Yao, Jean-Francois Ton, Xiaoying Zhang, Ruocheng Guo Hao Cheng, Yegor Klochkov, Muhammad Faaiz Taufiq, and Hang Li. Trustworthy llms: A survey and guideline for evaluating large language models’ alignment. arXiv preprint arXiv:2308.[05374, 2023](https://md.research.elamur.ai/free/en/papers/liu-2023-trustworthy-llms-survey-guideline-evaluating/).
34. Yifan Yao, Jinhao Duan, Kaidi Xu, Yuanfang Cai, Zhibo Sun, and Yue Zhang. [A survey on large language model (llm) security and privacy: The good, the bad, and the ugly](https://md.research.elamur.ai/free/en/papers/yao-2023-survey-large-language-model-llm/). High-Confidence Computing, page 100211, 2024.
35. Anthropic, Dec 2024. building-effective-agents. [find](https://scholar.google.com/scholar?q=Anthropic%2C+Dec+2024%2E+building%2Deffective%2Dagents%2E)
36. Ion Stoica, Matei Zaharia, Joseph Gonzalez, Ken Goldberg, Hao Zhang, Anastasios Angelopoulos, Shishir G Patil, Lingjiao Chen, Wei-Lin Chiang, and Jared Q Davis. Specifications: The missing link to making the development of llm systems an engineering discipline. arXiv preprint arXiv:2412.05299, 2024. [find](https://scholar.google.com/scholar?q=Ion+Stoica%2C+Matei+Zaharia%2C+Joseph+Gonzalez%2C+Ken+Goldberg%2C+Hao+Zhang%2C+Anastasios+Angelopoulos%2C+Shishir+G+Patil%2C+Lingjiao+Chen%2C+Wei%2DLin+Chiang%2C+and+Jared+Q+Davis%2E+Specifications%3A+The+missing+link+to+mak)
37. Gagan Bansal, Jennifer Wortman Vaughan, Saleema Amershi, Eric Horvitz, Adam Fourney, Hussein Mozannar, Victor Dibia, and Daniel S. Weld. Challenges in human-agent communication. Technical Report MSR-TR-2024-53, Microsoft, December 2024. URL [https://www.microsoft.com/en-us/research/publication/human-…](https://www.microsoft.com/en-us/research/publication/human-…) [↗](https://www.microsoft.com/en-us/research/publication/human-…)
38. Ge Bai, Jie Liu, Xingyuan Bu, Yancheng He, Jiaheng Liu, Zhanhui Zhou, Zhuoran Lin, Wenbo Su, Tiezheng Ge, Bo Zheng, and Wanli Ouyang. [Mt-bench-101: A fine-grained benchmark for evaluating large language models in multi-turn dialogues](https://md.research.elamur.ai/free/en/papers/bai-2024-mt-bench-101-fine-grained/). In Proceedings of the 62nd Annual Meeting of the Association for…
39. Song Da, Zijie Zhou, Zhijie Wang, Yuheng Huang, Shengmai Chen, Bonan Kou, Lei Ma, and Tianyi Zhang. An empirical study of code generation errors made by large language models. In In 7th Annual Symposium on Machine Programming, 2023. [find](https://scholar.google.com/scholar?q=Song+Da%2C+Zijie+Zhou%2C+Zhijie+Wang%2C+Yuheng+Huang%2C+Shengmai+Chen%2C+Bonan+Kou%2C+Lei+Ma%2C+and+Tianyi+Zhang%2E+An+empirical+study+of+code+generation+errors+made+by+large+language+models%2E+In+In+7th+Annual+Symposi)
40. Negar Arabzadeh, Siqing Huo, Nikhil Mehta, Qinqyun Wu, Chi Wang, Ahmed Awadallah, Charles L. A. Clarke, and Julia Kiseleva. Assessing and verifying task utility in llm-powered applications, 2024. URL https://arxiv.org/abs/2405.02178. [↗](https://arxiv.org/abs/2405.02178)
41. Will Epperson, Gagan Bansal, Victor Dibia, Adam Fourney, Jack Gerrits, Erkang (Eric) Zhu, and Saleema Amershi. Interactive debugging and steering of multi-agent ai systems. In CHI 2025, April 2025. URL https://arxiv.org/abs/2503.02068. — [arXiv:2503.02068](https://arxiv.org/abs/2503.02068) [↗](https://arxiv.org/abs/2503.02068)
42. Shaokun Zhang, Ming Yin, Jieyu Zhang, Jiale Liu, Zhiguang Han, Jingyang Zhang, Beibin Li, Chi Wang, Huazheng Wang, Yiran Chen, and Qingyun Wu. Which agent causes task failures and when? on automated failure attribution of llm multi-agent systems, 2025. URL https://arxiv.org/abs/2505.00212. URL https… — [arXiv:2505.00212](https://arxiv.org/abs/2505.00212) [↗](https://arxiv.org/abs/2505.00212)
43. Claire B Draucker, Donna S Martsolf, Ratchneewan Ross, and Thomas B Rusk. Theoretical sampling and category development in grounded theory. Qualitative health research, 17(8): 1137–1148, 2007. [find](https://scholar.google.com/scholar?q=Claire+B+Draucker%2C+Donna+S+Martsolf%2C+Ratchneewan+Ross%2C+and+Thomas+B+Rusk%2E+Theoretical+sampling+and+category+development+in+grounded+theory%2E+Qualitative+health+research%2C+17%288%29%3A+1137%E2%80%931148%2C+2007%2E)
44. Shahedul Huq Khandkar. Open coding. University of Calgary, 23(2009):2009, 2009. [find](https://scholar.google.com/scholar?q=Shahedul+Huq+Khandkar%2E+Open+coding%2E+University+of+Calgary%2C+23%282009%29%3A2009%2C+2009%2E)
45. Manus AI. Manus. [https://manus.im/](https://manus.im/), 2025. [↗](https://manus.im/)
46. Anthropic. Model context protocol: Introduction. [https://modelcontextprotocol.io/introduction](https://modelcontextprotocol.io/introduction), dec 2024. [↗](https://modelcontextprotocol.io/introduction)
47. Rao Surapaneni, Miku Jha, Michael Vakoc, and Todd Segal. A2a: A new era of agent interoperability, April 2025. URL [https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/). Google Developers Blog. [↗](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/)
48. Saaket Agashe, Yue Fan, Anthony Reyna, and Xin Eric Wang. Llm-coordination: Evaluating and analyzing multi-agent coordination abilities in large language models, 2025. URL https://arxiv.org/abs/2310.03903. [↗](https://arxiv.org/abs/2310.03903)
49. Charles Perrow. Normal Accidents: Living with High-Risk Technologies. Princeton University Press, Princeton, NJ, 1984. ISBN 978-0691004129. [find](https://scholar.google.com/scholar?q=Charles+Perrow%2E+Normal+Accidents%3A+Living+with+High%2DRisk+Technologies%2E+Princeton+University+Press%2C+Princeton%2C+NJ%2C+1984%2E+ISBN+978%2D0691004129%2E)
50. Karlene H. Roberts. New challenges in organizational research: High reliability organizations. Organization & Environment, 3(2):111–125, 1989. doi: 10.1177/108602668900300202. — [doi:10.1177/108602668900300202](https://doi.org/10.1177/108602668900300202) [find](https://scholar.google.com/scholar?q=Karlene+H%2E+Roberts%2E+New+challenges+in+organizational+research%3A+High+reliability+organizations%2E+Organization+%26+Environment%2C+3%282%29%3A111%E2%80%93125%2C+1989%2E+doi%3A+10%2E1177%2F108602668900300202%2E+%E2%80%94+%5Bdoi%3A10%2E1177%2F108602668)
51. Gene I Rochlin. Reliable organizations: Present research and future directions. Journal of contingencies and crisis management., 4(2), 1996. ISSN 0966-0879. [find](https://scholar.google.com/scholar?q=Gene+I+Rochlin%2E+Reliable+organizations%3A+Present+research+and+future+directions%2E+Journal+of+contingencies+and+crisis+management%2E%2C+4%282%29%2C+1996%2E+ISSN+0966%2D0879%2E)
52. Sirui Hong, Xiawu Zheng, Jonathan Chen, Yuheng Cheng, Jinlin Wang, Ceyao Zhang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, Liyang Zhou, et al. Metagpt: Meta programming for multi-agent collaborative framework. arXiv preprint arXiv:2308.[00352, 2023](https://md.research.elamur.ai/free/en/papers/hong-2023-metagpt-meta-programming-multi/).
53. Huy Nhat Phan, Tien N Nguyen, Phong X Nguyen, and Nghi DQ Bui. Hyperagent: Generalist software engineering agents to solve coding tasks at scale. arXiv preprint arXiv:2409.16299, 2024. — [arXiv:2409.16299](https://arxiv.org/abs/2409.16299) [find](https://scholar.google.com/scholar?q=Huy+Nhat+Phan%2C+Tien+N+Nguyen%2C+Phong+X+Nguyen%2C+and+Nghi+DQ+Bui%2E+Hyperagent%3A+Generalist+software+engineering+agents+to+solve+coding+tasks+at+scale%2E+arXiv+preprint+arXiv%3A2409%2E16299%2C+2024%2E+%E2%80%94+%5BarXiv%3A2409%2E1)
54. Harsh Trivedi, Tushar Khot, Mareike Hartmann, Ruskin Manku, Vinty Dong, Edward Li, Shashank Gupta, Ashish Sabharwal, and Niranjan Balasubramanian. [Appworld: A controllable world of apps and people for benchmarking interactive coding agents](https://md.research.elamur.ai/free/en/papers/trivedi-2024-appworld-controllable-world-apps-people/). arXiv preprint arXiv:2407.18901, 2024.
55. Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, Shaokun Zhang, Jiale Liu, et al. Autogen: Enabling next-gen llm applications via multi-agent conversations. In First Conference on Language Modeling, 2024. [find](https://scholar.google.com/scholar?q=Qingyun+Wu%2C+Gagan+Bansal%2C+Jieyu+Zhang%2C+Yiran+Wu%2C+Beibin+Li%2C+Erkang+Zhu%2C+Li+Jiang%2C+Xiaoyun+Zhang%2C+Shaokun+Zhang%2C+Jiale+Liu%2C+et+al%2E+Autogen%3A+Enabling+next%2Dgen+llm+applications+via+multi%2Dagent+conversati)
56. Chen Qian, Wei Liu, Hongzhang Liu, Nuo Chen, Yufan Dang, Jiahao Li, Cheng Yang, Weize Chen, Yusheng Su, Xin Cong, et al. [Chatdev: Communicative agents for software development](https://md.research.elamur.ai/free/en/papers/qian-2023-chatdev-communicative-agents-software/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1…
57. Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Shaokun Zhang, Erkang Zhu, Beibin Li, Li Jiang, Xiaoyun Zhang, and Chi Wang. [Autogen: Enabling next-gen llm applications via multi-agent conversation](https://md.research.elamur.ai/free/en/papers/wu-2023-autogen-enabling-next-gen-llm/) framework. arXiv preprint arXiv:2308.08155, 2023.
58. Jia He, Mukund Rungta, David Koleczek, Arshdeep Sekhon, Franklin X Wang, and Sadid Hasan. Does prompt formatting have any impact on llm performance? arXiv preprint arXiv:2411.10541, 2024. — [arXiv:2411.10541](https://arxiv.org/abs/2411.10541) [find](https://scholar.google.com/scholar?q=Jia+He%2C+Mukund+Rungta%2C+David+Koleczek%2C+Arshdeep+Sekhon%2C+Franklin+X+Wang%2C+and+Sadid+Hasan%2E+Does+prompt+formatting+have+any+impact+on+llm+performance%3F+arXiv+preprint+arXiv%3A2411%2E10541%2C+2024%2E+%E2%80%94+%5BarXiv%3A241)
59. Yashar Talebirad and Amirhossein Nadiri. [Multi-agent collaboration: Harnessing the power of intelligent llm agents](https://md.research.elamur.ai/free/en/papers/talebirad-2023-multi-agent-collaboration-harnessing-power/). arXiv preprint arXiv:2306.03314, 2023.
60. Chi-Min Chan, Weize Chen, Yusheng Su, Jianxuan Yu, Wei Xue, Shanghang Zhang, Jie Fu, and Zhiyuan Liu. [Chateval: Towards better llm-based evaluators through multi-agent debate](https://md.research.elamur.ai/free/en/papers/chan-2023-chateval-better-llm-based/). arXiv preprint arXiv:2308.07201, 2023.
61. Yixuan Weng, Minjun Zhu, Fei Xia, Bin Li, Shizhu He, Shengping Liu, Bin Sun, Kang Liu, and Jun Zhao. [Large language models are better reasoners with self-verification](https://md.research.elamur.ai/free/en/papers/weng-2023-large-language-models-are-better/). In The 2023 Conference on Empirical Methods in Natural Language Processing, 2023.
62. LangChain. Langgraph, 2024. URL [https://www.langchain.com/langgraph](https://www.langchain.com/langgraph). [↗](https://www.langchain.com/langgraph)
63. Anthropic. Building effective agents, 2024. URL [https://www.anthropic.com/research/building-effective-agents](https://www.anthropic.com/research/building-effective-agents). [↗](https://www.anthropic.com/research/building-effective-agents)
64. Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Tom Griffiths, Yuan Cao, and Karthik Narasimhan. Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems, 36, 2024. [find](https://scholar.google.com/scholar?q=Shunyu+Yao%2C+Dian+Yu%2C+Jeffrey+Zhao%2C+Izhak+Shafran%2C+Tom+Griffiths%2C+Yuan+Cao%2C+and+Karthik+Narasimhan%2E+Tree+of+thoughts%3A+Deliberate+problem+solving+with+large+language+models%2E+Advances+in+Neural+Informati)
65. Fatemeh Haji, Mazal Bethany, Maryam Tabar, Jason Chiang, Anthony Rios, and Peyman Najafirad. Improving llm reasoning with multi-agent tree-of-thought validator agent. arXiv preprint arXiv:2409.11527, 2024. [find](https://scholar.google.com/scholar?q=Fatemeh+Haji%2C+Mazal+Bethany%2C+Maryam+Tabar%2C+Jason+Chiang%2C+Anthony+Rios%2C+and+Peyman+Najafirad%2E+Improving+llm+reasoning+with+multi%2Dagent+tree%2Dof%2Dthought+validator+agent%2E+arXiv+preprint+arXiv%3A2409%2E11527%2C+)
66. Zhenran Xu, Senbao Shi, Baotian Hu, Jindi Yu, Dongfang Li, Min Zhang, and Yuxiang Wu. [Towards reasoning in large language models via multi-agent peer review collaboration](https://md.research.elamur.ai/free/en/papers/xu-2023-towards-reasoning-large-language-models/). arXiv preprint arXiv:2311.08152, 2023.
67. Benedikt Stroebl, Sayash Kapoor, and Arvind Narayanan. Inference scaling f laws: The limits of llm resampling with imperfect verifiers. arXiv preprint arXiv:2411.17501, 2024. [find](https://scholar.google.com/scholar?q=Benedikt+Stroebl%2C+Sayash+Kapoor%2C+and+Arvind+Narayanan%2E+Inference+scaling+f+laws%3A+The+limits+of+llm+resampling+with+imperfect+verifiers%2E+arXiv+preprint+arXiv%3A2411%2E17501%2C+2024%2E)
68. Lingjiao Chen, Jared Quincy Davis, Boris Hanin, Peter Bailis, Ion Stoica, Matei Zaharia, and James Zou. Are more llm calls all you need? towards scaling laws of compound inference systems. arXiv preprint arXiv:2403.02419, 2024. [find](https://scholar.google.com/scholar?q=Lingjiao+Chen%2C+Jared+Quincy+Davis%2C+Boris+Hanin%2C+Peter+Bailis%2C+Ion+Stoica%2C+Matei+Zaharia%2C+and+James+Zou%2E+Are+more+llm+calls+all+you+need%3F+towards+scaling+laws+of+compound+inference+systems%2E+arXiv+prepr)
69. Kush Jain, Gabriel Synnaeve, and Baptiste Rozière. Testgeneval: A real world unit test generation and test completion benchmark. arXiv preprint arXiv:2410.00752, 2024. — [arXiv:2410.00752](https://arxiv.org/abs/2410.00752) [find](https://scholar.google.com/scholar?q=Kush+Jain%2C+Gabriel+Synnaeve%2C+and+Baptiste+Rozie%CC%80re%2E+Testgeneval%3A+A+real+world+unit+test+generation+and+test+completion+benchmark%2E+arXiv+preprint+arXiv%3A2410%2E00752%2C+2024%2E+%E2%80%94+%5BarXiv%3A2410%2E00752%5D%28https%3A%2F%2Far)
70. Baolin Peng, Michel Galley, Pengcheng He, Hao Cheng, Yujia Xie, Yu Hu, Qiuyuan Huang, Lars Liden, Zhou Yu, Weizhu Chen, et al. [Check your facts and try again: Improving large language models with external knowledge and automated feedback](https://md.research.elamur.ai/free/en/papers/peng-2023-check-facts-try-again-improving/). arXiv preprint arXiv:2302.12813, 2023.
71. Pavan Kapanipathi, Ibrahim Abdelaziz, Srinivas Ravishankar, Salim Roukos, Alexander Gray, Ramon Astudillo, Maria Chang, Cristina Cornelio, Saswati Dana, Achille Fokoue, et al. Question answering over knowledge bases by leveraging semantic parsing and neuro-symbolic reasoning. arXiv preprint arXiv:20… — [Leveraging Abstract Meaning Representation for Knowledge Base Question Answering](https://md.research.elamur.ai/free/en/papers/kapanipathi-2021-leveraging-abstract-meaning-representation-knowledge/)
72. Xinyi Li, Sai Wang, Siqi Zeng, Yu Wu, and Yi Yang. A survey on llm-based multi-agent systems: workflow, infrastructure, and challenges. Vicinagearth, 1(1):9, 2024. [find](https://scholar.google.com/scholar?q=Xinyi+Li%2C+Sai+Wang%2C+Siqi+Zeng%2C+Yu+Wu%2C+and+Yi+Yang%2E+A+survey+on+llm%2Dbased+multi%2Dagent+systems%3A+workflow%2C+infrastructure%2C+and+challenges%2E+Vicinagearth%2C+1%281%29%3A9%2C+2024%2E)
73. Yaru Niu, Rohan R Paleja, and Matthew C Gombolay. Multi-agent graph-attention communication and teaming. In AAMAS, volume 21, page 20th, 2021. [find](https://scholar.google.com/scholar?q=Yaru+Niu%2C+Rohan+R+Paleja%2C+and+Matthew+C+Gombolay%2E+Multi%2Dagent+graph%2Dattention+communication+and+teaming%2E+In+AAMAS%2C+volume+21%2C+page+20th%2C+2021%2E)
74. Jiechuan Jiang and Zongqing Lu. Learning attentional communication for multi-agent cooperation. Advances in neural information processing systems, 31, 2018. [find](https://scholar.google.com/scholar?q=Jiechuan+Jiang+and+Zongqing+Lu%2E+Learning+attentional+communication+for+multi%2Dagent+cooperation%2E+Advances+in+neural+information+processing+systems%2C+31%2C+2018%2E)
75. Amanpreet Singh, Tushar Jain, and Sainbayar Sukhbaatar. Learning when to communicate at scale in multiagent cooperative and competitive tasks. arXiv preprint arXiv:1812.09755, 2018. — [arXiv:1812.09755](https://arxiv.org/abs/1812.09755) [find](https://scholar.google.com/scholar?q=Amanpreet+Singh%2C+Tushar+Jain%2C+and+Sainbayar+Sukhbaatar%2E+Learning+when+to+communicate+at+scale+in+multiagent+cooperative+and+competitive+tasks%2E+arXiv+preprint+arXiv%3A1812%2E09755%2C+2018%2E+%E2%80%94+%5BarXiv%3A1812%2E0975)
76. Chao Yu, Akash Velu, Eugene Vinitsky, Jiaxuan Gao, Yu Wang, Alexandre Bayen, and Yi Wu. The surprising effectiveness of ppo in cooperative multi-agent games. Advances in Neural Information Processing Systems, 35:24611–24624, 2022. [find](https://scholar.google.com/scholar?q=Chao+Yu%2C+Akash+Velu%2C+Eugene+Vinitsky%2C+Jiaxuan+Gao%2C+Yu+Wang%2C+Alexandre+Bayen%2C+and+Yi+Wu%2E+The+surprising+effectiveness+of+ppo+in+cooperative+multi%2Dagent+games%2E+Advances+in+Neural+Information+Processing+)
77. Xudong Guo, Daming Shi, Junjie Yu, and Wenhui Fan. Heterogeneous multi-agent reinforcement learning for zero-shot scalable collaboration. arXiv preprint arXiv:2404.03869, 2024. — [arXiv:2404.03869](https://arxiv.org/abs/2404.03869) [find](https://scholar.google.com/scholar?q=Xudong+Guo%2C+Daming+Shi%2C+Junjie+Yu%2C+and+Wenhui+Fan%2E+Heterogeneous+multi%2Dagent+reinforcement+learning+for+zero%2Dshot+scalable+collaboration%2E+arXiv+preprint+arXiv%3A2404%2E03869%2C+2024%2E+%E2%80%94+%5BarXiv%3A2404%2E03869%5D%28ht)
78. Weize Chen, Jiarui Yuan, Chen Qian, Cheng Yang, Zhiyuan Liu, and Maosong Sun. Optima: Optimizing effectiveness and efficiency for llm-based multi-agent system. arXiv preprint arXiv:2410.08115, 2024. [find](https://scholar.google.com/scholar?q=Weize+Chen%2C+Jiarui+Yuan%2C+Chen+Qian%2C+Cheng+Yang%2C+Zhiyuan+Liu%2C+and+Maosong+Sun%2E+Optima%3A+Optimizing+effectiveness+and+efficiency+for+llm%2Dbased+multi%2Dagent+system%2E+arXiv+preprint+arXiv%3A2410%2E08115%2C+2024%2E)
79. Eric Horvitz. Uncertainty, action, and interaction: In pursuit of mixed-initiative computing. IEEE Intelligent Systems, 14(5):17–20, 1999. [find](https://scholar.google.com/scholar?q=Eric+Horvitz%2E+Uncertainty%2C+action%2C+and+interaction%3A+In+pursuit+of+mixed%2Dinitiative+computing%2E+IEEE+Intelligent+Systems%2C+14%285%29%3A17%E2%80%9320%2C+1999%2E)
80. Barnali Chakraborty and Debapratim Purkayastha. Servicenow: From startup to world’s most innovative company. IUP Journal of Entrepreneurship Development, 20(1), 2023. [find](https://scholar.google.com/scholar?q=Barnali+Chakraborty+and+Debapratim+Purkayastha%2E+Servicenow%3A+From+startup+to+world%E2%80%99s+most+innovative+company%2E+IUP+Journal+of+Entrepreneurship+Development%2C+20%281%29%2C+2023%2E)
81. Qintong Li, Leyang Cui, Xueliang Zhao, Lingpeng Kong, and Wei Bi. [Gsm-plus: A comprehensive benchmark for evaluating the robustness of llms as mathematical problem solvers](https://md.research.elamur.ai/free/en/papers/li-2024-gsm-plus-comprehensive-benchmark-evaluating/). arXiv preprint arXiv:2402.19255, 2024.
82. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). arXiv preprint arXiv:2110.14168, 2021.
83. Binyuan Hui, Jian Yang, Zeyu Cui, Jiaxi Yang, Dayiheng Liu, Lei Zhang, Tianyu Liu, Jiajun Zhang, Bowen Yu, Kai Dang, et al. Qwen2. 5-coder technical report. arXiv preprint arXiv:2409.[12186, 2024](https://md.research.elamur.ai/free/en/papers/hui-2024-qwen2-5-coder-technical-report/).
84. Baptiste Roziere, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Romain Sauvestre, Tal Remez, et al. [Code llama: Open foundation models for code](https://md.research.elamur.ai/free/en/papers/roziere-2024-code-llama-open-foundation-models/). arXiv preprint arXiv:2308.12950, 2023. Organization of Appendix

*Updated 2 November 2026*