mlsanityData to 7 Oct 2026

Software engineering coding agents

2,041 papers32 this weekMedium confidence in this grouping: mostly one topic, with blurry edgesMedium confidence in this grouping: mostly one topic, with blurry edges · About

Keywords: code, software, agents, agent, coding

Key facts. Software engineering coding agents: 2,041 papers on arXiv since 2023 (data through 7 Oct 2026). 540 of its 2,041 mapped papers (26%) cite “SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering” (2024, arXiv 2405.15793). 312 of its 2,041 mapped papers (15%) cite “OpenHands: An Open Platform for AI Software Developers as Generalist Agents” (2024, arXiv 2407.16741). 201 new papers 17 Sep–7 Oct.

What’s new

17 Sep – 7 Oct 2026

201 new papers, 17 Sep – 7 Oct; 32 this week.

  1. This week · 32
  2. Cross-Provider Review as a Runtime Contract for Coding Agents: A Controlled Pilot and Fault-Injection Study
  3. The Harness as the Only Mutable Surface: Compliance-Bounded Self-Evolution of LLM Agents in Credit Pipelines, with a Measured Admission Gate also fits: Financial trading and forecasting agents
  4. Agent4RE: A Self-Refining Multi-agent Framework for End-to-End Software Requirements Engineering and Benchmarking
  5. TestJack: Should you trust the results in coding benchmarks? Agentic Coding Benchmarks Auditing via Evaluator Evolution
  6. Code Understanding is a Bottleneck for Coding Agents
  7. Beyond Type-checking: Towards Holistic Evaluation of Formal Specification Generation
  8. Why Software Engineering Is Indispensable in the Age of Coding Agents
  9. Know the Shape, Find the Fault: Topology-Conditioned Diagnosis of Multi-Agent LLM Failures also fits: Agent threats, guardrails and governance
  10. Cost-Efficient Theorem Proving via Agent Orchestration in Program Verification
  11. Coding-Agent Benchmarks Should Match Their Users' Task Flows
All 201
  1. We Query, Therefore We Compute: On Oracle Computation beyond the Machine, with an Application to Agents also fits: General LLM agent frameworks and benchmarks
  2. A Survey on LLM-Integrated Hardware Design Verification also fits: Tool learning and function calling
  3. GeoNatureAgent (GNA): A Framework and Benchmark for Pre-Production Evaluation of Tool-Using Agents on Geospatial and Environmental Tasks also fits: Tool learning and function calling
  4. RippleCP: Measuring Counterfactual Checkpoint Advantage in Coding Agents
  5. Large-scale Repository Engineering via Agent-Native Reusable Code Primitives
  6. Humanize: Judgement Engineering for Agentic Coding
  7. ScienceClaw: Benchmarking Continual Self-Evolution of AI-for-Science Agents Across the Natural and Social Sciences also fits: Agent skill learning and evolution
  8. ParanoiaEval: Benchmarking Unnecessary Defensive Work in Agentic Coding also fits: Agent threats, guardrails and governance
  9. A Case Study in Assuring AI-Written Software
  10. SquidAgent: Parallelize Wisely, Coordinate Efficiently also fits: General LLM agent frameworks and benchmarks
  11. Agentic RCA for Internet-Scale Services Using Constrained Creativity
  12. FC-SWE: Failure-Conditioned RL for Long-Horizon Software Engineering Agents
  13. OOPMAS: Object-Oriented Multi-Agent Systems for Query-Level Workflow Generation also fits: Web, search, and research agents
  14. Catching Developers in the Flow: Low-Latency Agentic Program Repair at Google Scale
  15. Verifying Coordination in Parallel Coding Agents: NP-Bench and a Scheduling Planner
  16. MemMux: Runtime Verification and Honest Resource Attribution for Fleets of Parallel Coding Agents
  17. Evaluating Inference Compute for Generative AI: A Framework for Enterprise Workloads also fits: Agent memory systems
  18. Back to the Future: Rethinking EDA Infrastructure for Agentic Systems in Chip Design Verification
  19. Can Agent Harnesses and Inference Engines Hear Each Other? The HEAR Protocol for Agentic LLM Serving
  20. Correct Code, Broken Contributions? SWE-CC: Benchmarking Repository Policy Compliance for Coding Agents
  21. Where Did the Repair First Go Wrong? Localizing the Origins of Silent Failures in Agentic Vulnerability Repair also fits: Agent threats, guardrails and governance
  22. Nexus: An Execution Fabric for AI Agents Across Cloud, Edge, and Devices also fits: Agent threats, guardrails and governance
  23. Earlier · 169
  24. CodeForge-MA: Execution-Verified Multi-Agent Learning with Language-Conditioned LoRA for Multilingual Code Generation
  25. MESH-Harness: Self-Improving Agent Harnesses via Bandit-Guided Compositional Evolution also fits: General LLM agent frameworks and benchmarks
  26. Fusion is the New Mutation: Bandit-Guided Evolution on Workflow Graphs
  27. StateWise: Diagnosing and Repairing Persistent Operational State Before Agent Actions
  28. Causal Improvement Graph for Agentic Harness Optimization also fits: Reinforcement learning for LLM agents
  29. SparseCraft: Agentic Hardware-Software Co-Optimization for Sparse Computing
  30. Software World Models: From Consequence Prediction to Decision Value
  31. Complex Agents, Shallow Tests: Demystifying and Enhancing Test Adequacy of Agent Harness in the Wild
  32. Viva La Vida: Verification and Accumulation Failures in Multi-Agent Proof Search also fits: General LLM agent frameworks and benchmarks
  33. AegisFlow: A Multi-Agent Agentic AI Framework for Autonomous Remediation and Self-Healing in Fragile Data Ecosystems
  34. Towards Automatically Pruning Logging Code with Coding Agents: How Far Are We?
  35. CRAFT: An Agentic Spreadsheet Form Filling System with Template Awareness
  36. Do Tool Calls Execute as Intended? Measuring and Repairing Intent-Execution Correspondence in LLM Agents also fits: Agent threats, guardrails and governance
  37. GAMEGO: Training Game-Dev Agents with Synthetic Trajectories Anchored in Real-World Assets also fits: Reinforcement learning for LLM agents
  38. Dynamic Harness Search: Building Multi-Agent Systems Per-Query via Prediction also fits: General LLM agent frameworks and benchmarks
  39. DUET: Co-Evolving Solver and Grader Agents also fits: General LLM agent frameworks and benchmarks
  40. Teaching Agents to Code Reliably
  41. Lightweight, Rubric-Guided Trajectory Evaluation for Production AI Agents
  42. D2K-Bench: Can LLM Agents Turn Expert Designs into Efficient GPU Kernels?
  43. hacktrace: behavior-supervised detection of reward hacking during code generation also fits: Agent threats, guardrails and governance
  44. GTDD: Generative Test-Driven Development for AI Coding Agents with Adversarial Testing
  45. When to Compile a Computer-Use Agent? Measuring Payback and Making Compilation Decisions for Token Efficiency
  46. VERSE: Verified Self-Evolving Optimizer for Agent Harnesses
  47. ProsaBuddy: Assisting Mechanized Real-Time Schedulability Analysis with LLM-based Agents
  48. SideKernel: A Usable microVM Sandbox for AI Coding Agents on macOS
  49. Inherit-MAS: Test-Time Evolution of Multi-Agent Systems through Workflow and Execution Inheritance also fits: General LLM agent frameworks and benchmarks
  50. Coco: An Agentic Copilot for the Hardware--Software Co-Design Lifecycle
  51. AutoCompact: Learning When to Compact Context in Long-Horizon Coding Agents
  52. TRACE: Tackling Real-World Resource Assignment Problems via Agentic Heuristic Design
  53. CONTRA: Discovering and Qualifying Behavior-Changing Questions for Selective Clarification in LLM Code Generation
  54. LLM-Driven Multi-Agent Control for Skill-Based Smart Manufacturing also fits: General LLM agent frameworks and benchmarks
  55. DeFA: Dependency-Guided Failure Attribution for LLM Agents also fits: Agent threats, guardrails and governance
  56. Auditing Action Settlement in LLM Agent Environments: Order, Progress, and Replay also fits: General LLM agent frameworks and benchmarks
  57. AgSpec: Pushing the Limits of Retrieval-Based Speculative Decoding in Coding Agent Pipelines also fits: Web, search, and research agents
  58. Groundability, Not Scale Alone: When Weak Reviewers Can Audit Strong Coding Agents
  59. Pay for the Fault, Not the Flow: Label-Free In-Flow Multi-Agent Workflow Optimization
  60. VeriHarness: Scaling Agentic Verification for Long-Horizon Tasks also fits: General LLM agent frameworks and benchmarks
  61. The Other Half of Workflow Portability: Evidence-Backed HPC Site Profiles with Agentic Discovery
  62. Understanding Issues, Causes and Solutions in Open-Source LLM-based Multi-Agent Systems
  63. FORALL-LEAN-AGENT for Auditable Reasoning in Formal Mathematics and Software Verification
  64. Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training also fits: Agent skill learning and evolution
  65. From Verification Failures to Reusable Guidance for Coding Agents
  66. Zero2Repo: Can Coding Agents Build Repositories from Scratch?
  67. Self-Evolving Coding Rules for AI Coding Agents also fits: General LLM agent frameworks and benchmarks
  68. Incident-Arena: Getting agents to the last nine of reliability
  69. Code That Works, Environments That Don't: Measuring Environment Reproducibility in AI-Generated Software
  70. How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering? also fits: General LLM agent frameworks and benchmarks
  71. Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents
  72. Learning When and How to Intervene: A Hindsight-Distilled Sentinel for Coding Agents also fits: Agent skill learning and evolution
  73. A2Z GameSpec-Bench: How Faithfully Can Coding Agents Generate Games from Game Design Specifications?
  74. Coding Agents for Coding Theory also fits: Web, search, and research agents
  75. SimEX: Simulation-Integrated Robotics AutoResearch also fits: Multimodal and embodied agents
  76. Composing Task-specific Agent Harnesses at Test Time with Reusable Primitives
  77. PatchHolmes: Agentic Patch Retrieval via Listwise Selection
  78. Staying on Task: Testing the Foundations of Long-Horizon Agent Reliability also fits: General LLM agent frameworks and benchmarks
  79. ASCEND: Personal AI Agents for Autonomous Scientific Computing Across HPC Clusters and GPU Workstations
  80. Rules to Tools: Executable Checks for LLM Agents in Scientific Computing
  81. E2E-SWE: Benchmarking LLMs on Building Working Codebases from Scratch
  82. MoFlow: Multi-Objective Agentic Workflow Generation also fits: General LLM agent frameworks and benchmarks
  83. AgentBug-Smith: Automatically Reproducing Real-World Harness Bugs in Agentic Systems
  84. LoLBench: Evaluating Coding Agents with Long-Horizon Proposals on Large Software Systems
  85. WitnessGym: Benchmarking Coding Agents on the Construction of Bug Witnesses
  86. MAADBench: The Refreshable Paradigm for Anomaly Detection in Multi-Agent Systems also fits: Agent threats, guardrails and governance
  87. Counterfactual Rollout Replay: Forkable Environments as Free Process Rewards for Software Engineering Agents also fits: Reinforcement learning for LLM agents
  88. Relic: From Multi-Agent Collaboration to Persistent Organizational Capability
  89. RepoMAS: Solving Progressively Specified Tasks with Issue-Driven Multi-Agent Systems
  90. LatentSift: Policy-State Filtering for Token-Efficient Verification of Software Engineering Agents
  91. From Dead Code and Static Requirements to Working Engines: Software Revival with Coding Agents
  92. An Exact Generate - Transform Decomposition of Small-LLM Team Scaling Across Orchestration Architectures
  93. Self-Evolving Coding Agents: From Digital Programs to Physical-World Intelligence also fits: Multimodal and embodied agents
  94. Planarian: Managing Agent State with Statepoints
  95. Do Coding Agents Reuse Existing Code or Reinvent the Wheel?
  96. Towards Reliable AI Data Scientists: Data Agents with Workflow Harnesses
  97. When Valid Tool Calls Change Meaning: Formation-Consistent Dispatch for LLM Agents also fits: Agent threats, guardrails and governance
  98. Audit the Scaffold, Not the Checkpoint: A Stationarity Dichotomy for Recursive Self-Improvement in Agentic Coding
  99. Codoku: Renewable Program-Reasoning Challenges for Frontier Coding Agents
  100. AgentWare: Automating the Lifecycle of Agentic Applications across the Edge-to-Cloud Continuum
  101. Certified Selective Automation of LLM Agent Evaluation also fits: Agent threats, guardrails and governance
  102. ControlScope: Workflow Revision and Reliability in LLM Agents
  103. StateGuard: Analytical-State Management with Validity-Aware Intervention for Long-Horizon Data Agents also fits: Agent threats, guardrails and governance
  104. Towards Certificate-Driven Software Porting: A Self-Improving Agentic Harness for Scientific Program Optimization
  105. Analyzing and Mitigating Cost-Inefficient Behaviors in Coding Agents
  106. Maat: Independent Deterministic Contract-Based Governance for Multi-Agent LLM Workflows also fits: Agent threats, guardrails and governance
  107. Opera: A Verbal Critic Framework for Long-horizon Coding Agents also fits: General LLM agent frameworks and benchmarks
  108. DynGraphAgentBench: A Benchmark for Agentic Lifecycle Control in Dynamic Graph Anomaly Detection also fits: Agent threats, guardrails and governance
  109. Vestrum: Improving Agent Harnesses by Adapting Their Verification, Structure and Memory also fits: Tool learning and function calling
  110. Identical Runs, Different Results: Benchmarking AI Coding Agents on Open-Weight Models also fits: General LLM agent frameworks and benchmarks
  111. EfficientAgent: What Makes KV Cache Offloading Work for Concurrent Agents?
  112. SWE-Game: Can Coding Agents Build the Games We Want?
  113. APEX: An Extensible Model for Agent-Assisted Production Scheduling
  114. Graph-Guided Repository Environment Construction
  115. TraceDance: An Automated System for Building Agent Behavior Benchmarks from Real-World Agent Deployment Traces
  116. LiteEvo: Automated, Cost-Efficient Harness Evolution for Generalization to Unseen Tasks also fits: General LLM agent frameworks and benchmarks
  117. Beyond Token Savings: A Systematic Study of Context Compression in LLM Agents also fits: Agent memory systems
  118. Re-derivability Decides What a Staged Agent Pipeline Recovers After an Upstream Fault also fits: Agent threats, guardrails and governance
  119. SWE-MILE: Asynchronous Potential-Induced Milestone Credit Assignment for Long-Horizon Software Engineering Agents
  120. CUA-SWE: When Computer-Use Agents Meet Visual Software Engineering
  121. Groupwise Agentic Grading and Advantage Redistribution for Code Agent RL also fits: Reinforcement learning for LLM agents
  122. Fail Loudly: An Auditable Runtime for Agentic Data Analysis also fits: Web, search, and research agents
  123. From Anomalies to Failures: Constructing Causal Error Graphs for Agentic Trace Diagnosis
  124. DAAF: From Failure Localization to Editable System Assets in LLM Agents
  125. PluginRSI: Recursive Improvement of Agent Harnesses with Reusable Plugins
  126. AuthorityLens: Rethinking LLM-Based Agent Systems Through the Lens of Authority also fits: Agent threats, guardrails and governance
  127. OptiArena: Can LLMs Improve Executable Algorithms under Fixed Resource Budgets? also fits: Tool learning and function calling
  128. Toward Interactive Understanding of Code APIs also fits: Tool learning and function calling
  129. Goal-Persistent Coding Agents as Scientific Performance Engineers: A Fixed-Radius Nearest-Neighbor Case Study
  130. Verification as an Architectural Layer for LLM Agents: A V-Model Design, and a Pilot Study of Its Deterministic Core
  131. Choir: An Open Protocol for Distributed Multi-Agent Autoformalization also fits: General LLM agent frameworks and benchmarks
  132. Omni-IO Skills: Harnessing Your Agent Omni-Native also fits: Agent skill learning and evolution
  133. Compact Documentation for Coding Agents: A Benchmark, an Optimizer, and Why It Does Not Transfer
  134. Compress What You See, Not What You Say: Anchored Context Distillation for Latent-Observation Software Engineering Agents also fits: Agent memory systems
  135. Beyond Approved Actions: Runtime Validation of Persistent Outcomes in Agent Workflows
  136. DynBranch: Speculative Subgraph Reuse for Dynamic Agentic LLM Serving
  137. Learning What to Skip: Counterfactual Credit Assignment for Efficient Multi-Agent LLM Workflows
  138. Synthesis Through Simulation: Generating Coherent Enterprise Data via Scalable Agent-System Interaction
  139. Evaluating Agent Skills for Version-Specific Plugin Migration: A Retrospective Study
  140. SWE-Prometheus: Measuring Engineering Governance Improvements in Real-World Repositories
  141. Where Does Exactly-Once Live? Model, Harness, and Tool-Contract Effects on Duplicate Side Effects in LLM Agents
  142. Control the Harness, Control the Cost: Routing and Governing AI Coding Agents in the Enterprise also fits: General LLM agent frameworks and benchmarks
  143. When Is a Multi-Agent Code Judge Actually Grounded? Two Label-Free Measurements, and a Judge That Declines to Guess also fits: Medical and scientific domain agents
  144. Don't Read the Log: Execution Traces Contaminate Verifiers in Video-Generation Agents
  145. From Agent Output to Authorized Transition
  146. Unity Insight: A Production Code--Asset Index for LLM Coding Agents in Unity Projects
  147. FDE-Bench: Evaluating LLM Agents for Deployment Environment Configuration
  148. WhatWorkedBench: Benchmarking Experimental Understanding in AI Agents also fits: Agent skill learning and evolution
  149. Specifying and Maintaining Agentic Workflows: An Empirical Study of GitHub Agentic Workflows
  150. From PyTorch to the NPU: LLM-Agent-Driven Model Conversion Across Heterogeneous Inference Runtimes also fits: General LLM agent frameworks and benchmarks
  151. TomasuLLM: Out-of-Order Speculative Execution for LLM Agents
  152. Crossflow: Prefill-Decode Elasticity for Agentic LLM Serving also fits: Financial trading and forecasting agents
  153. Harness as a Language: A Minimalist Agent Framework With Maximal Expressivity
  154. Who Finishes the Job? A Study of Follow-Up Fixes and Commit Authorship on AI Coding Agent Pull Requests
  155. CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents
  156. SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving
  157. Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents also fits: Agent skill learning and evolution
  158. From Approval to Execution: Reconstruction-Aware Repair Analysis for LLM-Agent Software
  159. FIRE: Failure-Informed Runtime Engineering for Reliable Language-Model Agents
  160. Evaluating Coding Agents on Kernel Exploit Generation
  161. Silent Failures in Agent-Tool Interaction: An Audit of ToolUniverse
  162. Spec2COBOLRot: An Agentic-AI Degradation Loop for Realistic COBOL Corpus Generation
  163. Passes Alone, Fails Together: Benchmarking Semantic Coordination in Parallel LLM-Agent Development
  164. Trains but Doesn't Learn: A Post-Training Delivery Benchmark for LLM Agents as Forward-Deployed Engineers
  165. Trajectory-Aware Benchmark Subset Selection for Cost-Efficient Software Engineering Agent Regression Testing
  166. A Lean and Spec-Driven AI-Assisted Software Development Lifecycle for Applied AI Education: The AI-SDLC Approach
  167. Canonical Procedural Actions: An Auditable Annotation Protocol for Tool-Use Agent Traces also fits: Tool learning and function calling
  168. The Work Behind Delegation: A Framework for Supervising AI Coding Agents
  169. APEXA: Execution-Integrity Enforcement for Multi-Agent LLM Automation of Synchrotron Data Reduction
  170. Self-Healing Harness for Runtime Oversight of Agent Self-Modification also fits: Agent threats, guardrails and governance
  171. EDGEGEN: Improving Tool-Calling Agents Beyond Happy Paths with Synthetic Edge Case Generation also fits: Tool learning and function calling
  172. What Makes a Terminal-Bench Task Hard? Separating Genuine Hardness from Fake-Hardness on an Adjudicated Agentic Corpus also fits: Agent skill learning and evolution
  173. Vibe-GUIDE: A Graph-based User Interface in IDEs for Oversight in Vibe Coding
  174. FLARE: A Full-Lifecycle Dense Supervision Paradigm for Long-Horizon Coding Agents via Generative Reward Model
  175. Total Cost of Agency: Exact Attribution of Memory Injection Cost in Multi-Agent LLM Workflows also fits: Tool learning and function calling
  176. VibeMemBench: Evaluating Memory Systems for Coding Agents on Real Repository Coding Tasks
  177. TicTacBench: Benchmarking Timing Closure Capabilities of Coding Agents
  178. CraftBench-UE: Deterministic Evaluation for Coding Agents in Unreal Engine
  179. LazyAgent: Demand-Driven Materialization and Physical Optimization of Agentic Programs
  180. AgentRouter: Heterogeneous Model Routing for Cost-Optimal Multi-Step Agentic Workflows also fits: General LLM agent frameworks and benchmarks
  181. From Code to Requirements: Agentic Reverse Engineering of Business Rules at Enterprise Scale
  182. CodeMidas: Scaling Agentic Coding RL Environments from Code Itself
  183. EnterpriseVal: Quantifying the Efficacy, Reliability and Value of Generative AI in the Enterprise also fits: General LLM agent frameworks and benchmarks
  184. GameLogicBench: Evaluating Coding Agents on Runtime Game Logic with Tick-Level State Assertions
  185. Efficient Benchmarking in Production: A Study of an Evolving LLM Agent
  186. Can Agents Design Better Chips with a Higher Level Abstraction?
  187. Quantifying Overclaiming Propensity in Frontier LLM Agents also fits: Agent threats, guardrails and governance
  188. An Empirical Study of Harness Design for Coding Agents
  189. Chronicle: Cut-Point Replay for Regression Testing of LLM Agents also fits: Agent memory systems
  190. SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness
  191. Spotlights: Discovering Improvement Opportunities in Software Repositories
  192. Self Improvement via Fast Tree-search also fits: Agent skill learning and evolution

Foundations

Most cited by papers here

  1. SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering2024cited by 540 here
  2. OpenHands: An Open Platform for AI Software Developers as Generalist Agents2024cited by 312 here
  3. AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation2023cited by 192 here
  4. ChatDev: Communicative Agents for Software Development2023cited by 154 here
  5. Agentless: Demystifying LLM-based Software Engineering Agents2024cited by 152 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 98 here
  2. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 89 here
  3. 2023WebArena: A Realistic Web Environment for Building Autonomous Agentscited by 117 here
  4. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 316 here
  5. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 192 here
  6. 2023A Survey on Large Language Model based Autonomous Agentscited by 161 here
  7. 2023The Rise and Potential of Large Language Model Based Agents: A Surveycited by 94 here
  8. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 540 here
  9. 2024Agentless: Demystifying LLM-based Software Engineering Agentscited by 152 here
  10. 2024OpenHands: An Open Platform for AI Software Developers as Generalist Agentscited by 312 here
  11. 2024Training Software Engineering Agents and Verifiers with SWE-Gymcited by 91 here
  12. 2025SWE-smith: Scaling Data for Software Engineering Agentscited by 81 here
12 more it builds on, not in this chain
  1. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 219 here
  2. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 278 here
  3. 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 113 here
  4. 2023ChatDev: Communicative Agents for Software Developmentcited by 154 here
  5. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 107 here
  6. 2023AgentBench: Evaluating LLMs as Agentscited by 94 here
  7. 2023AgentCoder: Multi-Agent-based Code Generation with Iterative Testing and Optimisationcited by 95 here
  8. 2024Executable Code Actions Elicit Better LLM Agentscited by 93 here
  9. 2024RepairAgent: An Autonomous, LLM-Based Agent for Program Repaircited by 102 here
  10. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 110 here
  11. 2024Large Language Model-Based Agents for Software Engineering: A Surveycited by 78 here
  12. 2025Why Do Multi-Agent LLM Systems Fail?cited by 119 here

The 24 most cited of 2,963.

Browse

30 of 2,041, ranked 6–35

  1. 6Why Do Multi-Agent LLM Systems Fail?2025119
  2. 7RepairAgent: An Autonomous, LLM-Based Agent for Program Repair2024102
  3. 8AgentCoder: Multi-Agent-based Code Generation with Iterative Testing and Optimisation202395
  4. 9Executable Code Actions Elicit Better LLM Agents202493
  5. 10Training Software Engineering Agents and Verifiers with SWE-Gym202491
  6. 11SWE-smith: Scaling Data for Software Engineering Agents202581
  7. 12Large Language Model-Based Agents for Software Engineering: A Survey202478
  8. 13CodeAgent: Enhancing Code Generation with Tool-Integrated Agent Systems for Real-World Repo-level Coding Challenges202467
  9. 14AFlow: Automating Agentic Workflow Generation202467
  10. 15Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems202560
All 30
  1. 16The Rise of AI Teammates in Software Engineering (SE) 3.0: How Autonomous Coding Agents Are Reshaping Software Engineering202557
  2. 17Understanding Software Engineering Agents: A Study of Thought-Action-Result Trajectories202555
  3. 18SWE-rebench: An Automated Pipeline for Task Collection and Decontaminated Evaluation of Software Engineering Agents202554
  4. 19MAGIS: LLM-Based Multi-Agent Framework for GitHub Issue Resolution202454
  5. 20SWT-Bench: Testing and Validating Real-World Bug-Fixes with Code Agents202454
  6. 21SWE-Search: Enhancing Software Agents with Monte Carlo Tree Search and Iterative Refinement202452
  7. 22Self-collaboration Code Generation via ChatGPT202352
  8. 23LLM-Based Multi-Agent Systems for Software Engineering: Literature Review, Vision and the Road Ahead202451
  9. 24Trae Agent: An LLM-based Agent for Software Engineering with Test-time Scaling202550
  10. 25Live-SWE-agent: Can Software Engineering Agents Self-Evolve on the Fly?202547
  11. 26MapCoder: Multi-Agent Code Generation for Competitive Problem Solving202447
  12. 27AlphaEvolve: A coding agent for scientific and algorithmic discovery202546
  13. 28AgenTracer: Who Is Inducing Failure in the LLM Agentic Systems?202546
  14. 29TRAIL: Trace Reasoning and Agentic Issue Localization202543
  15. 30R2E-Gym: Procedural Environments and Hybrid Verifiers for Scaling Open-Weights SWE Agents202541
  16. 31On the Use of Agentic Coding: An Empirical Study of Pull Requests on GitHub202541
  17. 32Where LLM Agents Fail and How They can Learn From Failures202541
  18. 33SpecRover: Code Intent Extraction via LLMs202439
  19. 34A Survey on Code Generation with LLM-based Agents202534
  20. 35MarsCode Agent: AI-native Automated Bug Fixing202434