AI Coding Agents

Follow this topic Ask about this topic

AI coding agents study how language models can work with repositories and development tools to carry out tasks such as fixing bugs, adding features, and building software. Common approaches combine planning and tool use with repository search, structured code editing, reusable skills or memory, and feedback from tests or execution. Benchmarks such as SWE-bench and Terminal-Bench measure issue resolution and command-line work, while newer evaluations test long-running, multi-file changes and behavior in real applications. Agents can solve many coding tasks, but success varies with the model, harness, and task; passing functional tests still does not guarantee secure, maintainable, or specification-faithful software, so reliable verification and effective human oversight remain open challenges.

Latest papers

329 papers62 papers in the last four weeks, up 130% on the four weeks before. 0.4% of all new papers.

Apr 21, 2026

AblateCell: A Reproduce-then-Ablate Agent for Virtual Cell Repositories

Systematic ablations are essential to attribute performance gains in AI Virtual Cells, yet they are rarely performed because biological repositories are under-standardized and tightly coupled to domain-specific data and formats. While recent coding agents can translate ideas into implementations, they typically stop at producing code and lack a verifier that can reproduce strong baselines and rigorously test which components truly matter. We introduce AblateCell, a reproduce-then-ablate agent for virtual cell repositories that closes this verification gap. AblateCell first reproduces reported baselines end-to-end by auto-configuring environments, resolving dependency and data issues, and rerunning official evaluations while emitting verifiable artifacts. It then conducts closed-loop ablation by generating a graph of isolated repository mutations and adaptively selecting experiments under a reward that trades off performance impact and execution cost. Evaluated on three single-cell perturbation prediction repositories (CPA, GEARS, BioLORD), AblateCell achieves 88.9% (+29.9% to human expert) end-to-end workflow success and 93.3% (+53.3% to heuristic) accuracy in recovering ground-truth critical components. These results enable scalable, repository-grounded verification and attribution directly on biological codebases.

Xue Xia, Chengkai Yao, Mingyu Tsoi +10

Shanghai Artificial Intelligence Laboratory · The Hong Kong University of Science and Technology (Guangzhou) · University of California San Diego +4

AI Coding Agents ML Reproducibility

Apr 18, 2026

HiveMind: OS-Inspired Scheduling for Concurrent LLM Agent Workloads

When multiple LLM coding agents share a rate-limited API endpoint, they exhibit resource contention patterns analogous to unscheduled OS processes competing for CPU, memory, and I/O. In a motivating incident, 3 of 11 parallel agents died from connection resets and HTTP 502 errors - a 27% failure rate - despite the API having sufficient aggregate capacity to serve all 11 sequentially. We present HIVEMIND, a transparent HTTP proxy that applies five OS-inspired scheduling primitives - admission control, rate-limit tracking, AIMD backpressure with circuit breaking, token budget management, and priority queuing - to eliminate the failure modes caused by uncoordinated parallel execution. The proxy requires zero modifications to existing agent code and supports Anthropic, OpenAI, and local model APIs via auto-detected provider profiles. Our evaluation across seven scenarios (5-50 concurrent agents) shows that uncoordinated agents fail at 72-100% rates under contention, while HIVEMIND reduces failures to 0-18% and eliminates 48-100% of wasted compute. An ablation study reveals that transparent retry - not admission control - is the single most critical primitive, but the primitives are most effective in combination. Real-world validation against Ollama confirms that HIVEMIND adds under 3ms of proxy overhead per request. The system is open-source under the MIT license.

Justice Owusu Agyemang, Jerry John Kponyo, Obed Kwasi Somuah +3

Sperix Labs · VIA Cybersecurity Lab, KNUST · Quantum and Assistive Technologies Lab, KNUST

LLM Inference Efficiency AI Coding Agents LLM Agent Orchestration LLM Agent Reliability LLM Request Scheduling

Apr 18, 2026

From Clinical Intent to Clinical Model: Autonomous Coding-Agents for Clinician-driven AI Development

Developing AI models that are useful in clinical practice, requires efficient collaboration between clinicians and AI developers. This poses a practical challenge: clinicians must repeatedly communicate and refine their requirements with AI developers before those requirements can be translated into executable model development. This iterative process is time-consuming, and even after repeated discussion, misalignment may still exist because the two sides do not fully share each other's expertise. Coding agents may help close this gap. They can write and refine code on their own, and they carry working knowledge of both medicine and AI to understand commands formulated by both medical experts and developers. We present a prototype that lets clinicians drive AI development directly. A clinician describes the task in plain language, and the system turns the description into a working pipeline, refines it through repeated experiments together with the clinician, and returns a model that meets the stated clinical objective. Across five clinical tasks, the system reliably produces models that matched the clinician's request and reached competitive performance. Most notably, on chest radiographs the system sharply reduced the model's reliance on chest drains, a well-known shortcut for pneumothorax classification, from 60% to 31% on one dataset and from 50% to 18% on another. Our results suggest that coding agents can shift clinical AI development toward a more clinician-driven mode, allowing domain experts to shape models directly instead of relaying requirements through specialized AI teams.

Zihao Zhao, Frederik Hauke, Juliana De Castilhos +4

University Hospital Aachen, Department of Diagnostic and Interventional Radiology, 52074 Aachen, Germany · Juelich AI Factory, Forschungszentrum Jülich, Jülich, Germany · Else Kröner Fresenius Center for Digital Health, TU Dresden, Dresden, Germany +2

AI Coding Agents Agentic Code Generation Human-AI Collaboration

Apr 16, 2026

Scaling Test-Time Compute for Agentic Coding

Test-time scaling has become a powerful way to improve large language models. However, existing methods are best suited to short, bounded outputs that can be directly compared, ranked or refined. Long-horizon coding agents violate this premise: each attempt produces an extended trajectory of actions, observations, errors, and partial progress taken by the agent. In this setting, the main challenge is no longer generating more attempts, but representing prior experience in a form that can be effectively selected from and reused. We propose a test-time scaling framework for agentic coding based on compact representations of rollout trajectories. Our framework converts each rollout into a structured summary that preserves its salient hypotheses, progress, and failure modes while discarding low-signal trace details. This representation enables two complementary forms of inference-time scaling. For parallel scaling, we introduce Recursive Tournament Voting (RTV), which recursively narrows a population of rollout summaries through small-group comparisons. For sequential scaling, we adapt Parallel-Distill-Refine (PDR) to the agentic setting by conditioning new rollouts on summaries distilled from prior attempts. Our method consistently improves the performance of frontier coding agents across SWE-Bench Verified and Terminal-Bench v2.0. For example, by using our method Claude-4.5-Opus improves from 70.9% to 77.6% on SWE-Bench Verified (mini-SWE-agent) and 46.9% to 59.1% on Terminal-Bench v2.0 (Terminus 1). Our results suggest that test-time scaling for long-horizon agents is fundamentally a problem of representation, selection, and reuse.

Joongwon Kim, Wannan Yang, Kelvin Niu +13

1Meta Superintelligence Labs · University of Washington · 3New York University +4

AI Coding Agents Test-Time Scaling Long-Horizon LLM Agents

Apr 14, 2026

Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems

Claude Code is an agentic coding tool that can run shell commands, edit files, and call external services on behalf of the user. This study describes its architecture by analyzing the publicly available source code and comparing it with two independent open-source AI agent systems, OpenClaw and Hermes Agent, that answer many of similar or even the same design questions. Our analysis identifies five human values, philosophies, and needs that motivate the architecture: human decision authority, safety, security, and privacy, reliable execution, capability amplification, and contextual adaptability. We then trace them through thirteen design principles to implementation choices. The core of the system is a simple while-loop that calls the model, runs tools, and repeats. Most of the code, however, lives in the systems around this loop: a permission system with seven modes and an ML-based classifier, a five-layer compaction pipeline for context management, four extensibility mechanisms (MCP, plugins, skills, and hooks), a subagent delegation and orchestration mechanism, and append-oriented session storage. Comparisons with OpenClaw and Hermes Agent show that the same design questions produce different answers across three deployment contexts. Claude Code emphasizes per-action safety, OpenClaw emphasizes perimeter-level access, and Hermes renders per-action approvals across many surfaces. At the runtime layer, Claude Code uses a single CLI loop, OpenClaw embeds the runtime within a gateway control plane, and Hermes uses one process whose role is set by its entry point. At the context and extension layer, Claude Code extends the context window, OpenClaw registers gateway-wide capabilities, and Hermes provides pluggable memory and model backends. We finally identify six open design directions for future agent systems, grounded in recent empirical, architectural, and policy literature.

Jiacheng Liu, Xiaohan Zhao, Xinyi Shang +1

VILA Lab, Mohamed bin Zayed University of Artificial Intelligence · University College London

AI Coding Agents LLM Agent Orchestration LLM Agent Security AI Agent Architecture LLM Agent Safety

Apr 14, 2026

Local-Splitter: A Measurement Study of Seven Tactics for Reducing Cloud LLM Token Usage on Coding-Agent Workloads

We present a systematic measurement study of seven tactics for reducing cloud LLM token usage when a small local model can act as a triage layer in front of a frontier cloud model. The tactics are: (1) local routing, (2) prompt compression, (3) semantic caching, (4) local drafting with cloud review, (5) minimal-diff edits, (6) structured intent extraction, and (7) batching with vendor prompt caching. We implement all seven in an open-source shim that speaks both MCP and the OpenAI-compatible HTTP surface, supporting any local model via Ollama and any cloud model via an OpenAI-compatible endpoint. We evaluate each tactic individually, in pairs, and in a greedy-additive subset across four coding-agent workload classes (edit-heavy, explanation-heavy, general chat, RAG-heavy). We measure tokens saved, dollar cost, latency, and routing accuracy. Our headline finding is that T1 (local routing) combined with T2 (prompt compression) achieves 45-79% cloud token savings on edit-heavy and explanation-heavy workloads, while on RAG-heavy workloads the full tactic set including T4 (draft-review) achieves 51% savings. We observe that the optimal tactic subset is workload-dependent, which we believe is the most actionable finding for practitioners deploying coding agents today.

Justice Owusu Agyemang, Jerry John Kponyo, Elliot Amponsah +2

Sperix Labs · VIA Cybersecurity Lab, KNUST · Quantum and Assistive Technologies Lab, KNUST

LLM Agent Workflow Optimization LLM Inference Efficiency LLM Routing AI Coding Agents Prompt Compression Semantic Caching

Apr 13, 2026

CodeTracer: Towards Traceable Agent States

Code agents are advancing rapidly, but debugging them is becoming increasingly difficult. As frameworks orchestrate parallel tool calls and multi-stage workflows over complex tasks, making the agent's state transitions and error propagation hard to observe. In these runs, an early misstep can trap the agent in unproductive loops or even cascade into fundamental errors, forming hidden error chains that make it hard to tell when the agent goes off track and why. Existing agent tracing analyses either focus on simple interaction or rely on small-scale manual inspection, which limits their scalability and usefulness for real coding workflows. We present CodeTracer, a tracing architecture that parses heterogeneous run artifacts through evolving extractors, reconstructs the full state transition history as a hierarchical trace tree with persistent memory, and performs failure onset localization to pinpoint the failure origin and its downstream chain. To enable systematic evaluation, we construct CodeTraceBench from a large collection of executed trajectories generated by four widely used code agent frameworks on diverse code tasks (e.g., bug fixing, refactoring, and terminal interaction), with supervision at both the stage and step levels for failure localization. Experiments show that CodeTracer substantially outperforms direct prompting and lightweight baselines, and that replaying its diagnostic signals consistently recovers originally failed runs under matched budgets. Our code and data are publicly available.

Han Li, Yifan Yao, Letian Zhu +13

Nanjing University · Alibaba Group · Renmin University of China +3

AI Coding Agents Agent Trajectory Analysis Agent Failure Analysis

Apr 13, 2026

From Translation to Superset: Benchmark-Driven Evolution of a Production AI Agent from Rust to Python

Cross-language migration of large software systems is a persistent engineering challenge, particularly when the source codebase evolves rapidly. We present a methodology for LLM-assisted continuous code translation in which a large language model translates a production Rust codebase (648K LOC, 65 crates) into Python (41K LOC, 28 modules), with public agent benchmarks as the objective function driving iterative refinement. Our subject system is Codex CLI, a production AI coding agent. We demonstrate that: (1) the Python port resolves 59/80 SWE-bench Verified tasks (73.8%) versus Rust's 56/80 (70.0%), and achieves 42.5% on Terminal-Bench versus Rust's 47.5%, confirming near-parity on real-world agentic tasks; (2) benchmark-driven debugging, revealing API protocol mismatches, environment pollution, a silent WebSocket failure mode, and an API 400 crash, is more effective than static testing alone; (3) the architecture supports continuous upstream synchronisation via an LLM-assisted diff-translate-test loop; and (4) the Python port has evolved into a capability superset with 30 feature-flagged extensions (multi-agent orchestration, semantic memory, guardian safety, cost tracking) absent from Rust, while preserving strict parity mode for comparison. Our evaluation shows that for LLM-based agents where API latency dominates, Python's expressiveness yields a 15.9x code reduction with negligible performance cost, while the benchmark-as-objective-function methodology provides a principled framework for growing a cross-language port from parity into an extended platform.

Jinhua Wang, Biswa Sengupta

LLM Suite Team, JP Morgan Chase & Co.

Code Translation AI Coding Agents

Apr 12, 2026

Resilient Write: A Six-Layer Durable Write Surface for LLM Coding Agents

LLM-powered coding agents increasingly rely on tool-use protocols such as the Model Context Protocol (MCP) to read and write files on a developer's workstation. When a write fails - due to content filters, truncation, or an interrupted session - the agent typically receives no structured signal, loses the draft, and wastes tokens retrying blindly. We present Resilient Write, an MCP server that interposes a six-layer durable write surface between the agent and the filesystem. The layers - pre-flight risk scoring, transactional atomic writes, resume-safe chunking, structured typed errors, out-of-band scratchpad storage, and task-continuity handoff envelopes - are orthogonal and independently adoptable. Each layer maps to a concrete failure mode observed during a real agent session in April 2026, in which content-safety filters silently rejected a draft containing redacted API-key prefixes. Three additional tools - chunk preview, format-aware validation, and journal analytics - emerged from using the system to compose this paper. A 186-test suite validates correctness at each layer, and quantitative comparison against naive and defensive baselines shows a 5x reduction in recovery time and a 13x improvement in agent self-correction rate. Resilient Write is open-source under the MIT license.

Justice Owusu Agyemang, Jerry John Kponyo, Elliot Amponsah +2

Sperixlabs, Ghana · Kwame Nkrumah University of Science and Technology, Kumasi, Ghana · VIA Cybersecurity Lab, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana

LLM Agent Reliability AI Coding Agents LLM Tool Use LLM Agent Harnesses Model Context Protocol

Apr 11, 2026

Applying an Agentic Coding Tool for Improving Published Algorithm Implementations

We present a two-stage pipeline for AI-assisted improvement of published algorithm implementations. In the first stage, a large language model with research capabilities identifies recently published algorithms satisfying explicit experimental criteria. In the second stage, Claude Code is given a prompt to reproduce the reported baseline and then iterate an improvement process. We apply this pipeline to published algorithm implementations spanning multiple research domains. Claude Code reported that all eleven experiments yielded improvements. Each improvement could be achieved within a single working day. We analyse the human contributions that remain indispensable, including selecting the target, verifying experimental validity, assessing novelty and impact, providing computational resources, and writing with appropriate AI-use disclosure. Finally, we discuss implications for peer review and academic publishing.

Worasait Suwannik

AI-Assisted Software Development AI Coding Agents

Apr 11, 2026

Formal Architecture Descriptors as Navigation Primitives for AI Coding Agents

AI coding agents spend a substantial fraction of their tool calls on undirected codebase exploration. We investigate whether providing agents with formal architecture descriptors can reduce this navigational overhead. We present three complementary studies. First, a controlled experiment (24 code localization tasks x 4 conditions, Claude Sonnet 4.6, temperature=0) demonstrates that architecture context reduces navigation steps by 33-44% (Wilcoxon p=0.009, Cohen's d=0.92), with no significant format difference detected across S-expression, JSON, YAML, and Markdown. Second, an artifact-vs-process experiment (15 tasks x 3 conditions) demonstrates that an automatically generated descriptor achieves 100% accuracy versus 80% blind (p=0.002, d=1.04), proving direct navigational value independent of developer self-clarification. Third, an observational field study across 7,012 Claude Code sessions shows 52% reduction in agent behavioral variance. A writer-side experiment (96 generation runs, 96 error injections) reveals critical failure mode differences: JSON fails atomically, YAML silently corrupts 50% of errors, S-expressions detect all structural completeness errors. We propose intent.lisp, an S-expression architecture descriptor, and open-source the Forge toolkit.

Ruoqi Jin

Independent Researcher, China

Code Localization AI Coding Agents Software Engineering

Apr 7, 2026

Beyond Functional Correctness: Design Issues in AI IDE-Generated Large-Scale Projects

New generation of AI coding tools, including AI-powered IDEs equipped with agentic capabilities, can generate code within the context of the project. These AI IDEs are increasingly perceived as capable of producing project-level code at scale. However, there is limited empirical evidence on the extent to which they can generate large-scale software systems and what design issues such systems may exhibit. To address this gap, we conducted a study to explore the capability of Cursor in generating large-scale projects and to evaluate the design quality of projects generated by Cursor. First, we propose a Feature-Driven Human-In-The-Loop (FD-HITL) framework that systematically guides project generation from curated project descriptions. We generated 10 projects using Cursor with the FD-HITL framework across three application domains and multiple technologies. We assessed the functional correctness of these projects through manual evaluation, obtaining an average functional correctness score of 91%. Next, we analyzed the generated projects using two static analysis tools, CodeScene and SonarQube, to detect design issues. We identified 1,305 design issues categorized into 9 categories by CodeScene and 3,193 issues in 11 categories by SonarQube. Our findings show that (1) when used with the FD-HITL framework, Cursor can generate functional large-scale projects averaging 16,965 LoC and 114 files; (2) the generated projects nevertheless contain design issues that may pose long-term maintainability and evolvability risks, requiring careful review by experienced developers; (3) the most prevalent issues include Code Duplication, high Code Complexity, Large Methods, Framework Best-Practice Violations, Exception-Handling Issues and Accessibility Issues; (4) these design issues violate design principles such as SRP, SoC, and DRY. The replication package is at https://github.com/Kashifraz/DIinAGP

Syed Mohammad Kashif, Ruiyin Li, Peng Liang +4

School of Computer Science, Wuhan University, China · School of Mathematical and Computational Sciences, Massey University, New Zealand · School of Computer Science, Nanjing University of Science and Technology, China +2

Software Engineering AI Coding Agents Repository-Level Code Generation Code Generation Evaluation

Apr 7, 2026

CODESTRUCT: Code Agents over Structured Action Spaces

LLM-based code agents treat repositories as unstructured text, applying edits through brittle string matching that frequently fails due to formatting drift or ambiguous patterns. We propose reframing the codebase as a structured action space where agents operate on named AST entities rather than text spans. Our framework, CODESTRUCT, provides readCode for retrieving complete syntactic units and editCode for applying syntax-validated transformations to semantic program elements. Evaluated on SWE-Bench Verified across six LLMs, CODESTRUCT improves Pass@1 accuracy by 1.2-5.0% while reducing token consumption by 12-38% for most models. Models that frequently fail to produce valid patches under text-based interfaces benefit most: GPT-5-nano improves by 20.8% as empty-patch failures drop from 46.6% to 7.2%. On CodeAssistBench, we observe consistent accuracy gains (+0.8-4.4%) with cost reductions up to 33%. Our results show that structure-aware interfaces offer a more reliable foundation for code agents.

Myeongsoo Kim, Joe Hsu, Dingmin Wang +3

AWS AI Labs

AI Coding Agents Code Generation

Apr 7, 2026

Spec Kit Agents: Context-Grounded Agentic Workflows

Spec-driven development (SDD) with AI coding agents provides a structured workflow, but agents often remain "context blind" in large, evolving repositories, leading to hallucinated APIs and architectural violations. We present Spec Kit Agents, a multi-agent SDD pipeline (with PM and developer roles) that adds phase-level, context-grounding hooks. Read-only probing hooks ground each stage (Specify, Plan, Tasks, Implement) in repository evidence, while validation hooks check intermediate artifacts against the environment. We evaluate 128 runs covering 32 features across five repositories. Context-grounding hooks improve judged quality by +0.15 on a 1-5 composite LLM-as-judge score (+3.0 percent of the full score; Wilcoxon signed-rank, p < 0.05) while maintaining 99.7-100 percent repository-level test compatibility. We further evaluate the framework on SWE-bench Lite, where augmentation hooks improve baseline by 1.7 percent, achieving 58.2 percent Pass@1.

Pardis Taghavi, Santosh Bhavani

Software Engineering AI Coding Agents Agentic Code Generation LLM Agent Workflow Optimization

Apr 5, 2026

Architecture Without Architects: How AI Coding Agents Shape Software Architecture

AI coding agents select frameworks, scaffold infrastructure, and wire integrations, often in seconds. These are architectural decisions, yet almost no one reviews them as such. We identify five mechanisms by which agents make implicit architectural choices and propose six prompt-architecture coupling patterns that map natural-language prompt features to the infrastructure they require. The patterns range from contingent couplings (structured output validation) that may weaken as models improve to fundamental ones (tool-call orchestration) that persist regardless of model capability. An illustrative demonstration confirms that prompt wording alone produces structurally different systems for the same task. We term the phenomenon vibe architecting, architecture shaped by prompts rather than deliberate design, and outline review practices, decision records, and tooling to bring these hidden decisions under governance.

Phongsakon Mark Konrad, Tim Lukas Adam, Riccardo Terrenzi +1

Centre for Industrial Software, University of Southern Denmark, Alsion 2, Sønderborg, 6400, Denmark

Software Architecture Design AI Coding Agents Software Engineering AI Agent Architecture AI Agent Governance Prompt Engineering

Apr 4, 2026

Squeez: Task-Conditioned Tool-Output Pruning for Coding Agents

Coding agents repeatedly consume long tool observations even though only a small fraction of each observation matters for the next step. We study task-conditioned tool-output pruning: given a focused query and one tool output, return the smallest verbatim evidence block the agent should inspect next. We introduce a benchmark of 11,477 examples built from SWE-bench repository interactions and synthetic multi-ecosystem tool outputs, with a manually curated 618-example test set. We fine-tune Qwen 3.5 2B with LoRA and compare it against larger zero-shot models and heuristic pruning baselines. Our model reaches 0.86 recall and 0.80 F1 while removing 92% of input tokens, outperforming zero-shot Qwen 3.5 35B A3B by 11 recall points and all heuristic baselines by a wide margin.

Ádám Kovács

KR Labs

AI Coding Agents LLM Agent Training

Apr 4, 2026

Measuring the Permission Gate: A Stress-Test Evaluation of Claude Code's Auto Mode

Claude Code's auto mode is the first deployed permission system for AI coding agents, using a two-stage transcript classifier to gate dangerous tool calls. Anthropic reports a 0.4% false positive rate and 17% false negative rate on production traffic. We present the first independent evaluation of this system on deliberately ambiguous authorization scenarios, i.e., tasks where the user's intent is clear but the target scope, blast radius, or risk level is underspecified. Using AmPermBench, a 128-prompt benchmark spanning four DevOps task families and three controlled ambiguity axes, we evaluate 253 state-changing actions at the individual action level against oracle ground truth. Our findings characterize auto mode's scope-escalation coverage under this stress-test workload. The end-to-end false negative rate is 81.0% (95% CI: 73.8%-87.4%), substantially higher than the 17% reported on production traffic, reflecting a fundamentally different workload rather than a contradiction. Notably, 36.8% of all state-changing actions fall outside the classifier's scope via Tier 2 (in-project file edits), contributing to the elevated end-to-end FNR. Even restricting to the 160 actions the classifier actually evaluates (Tier 3), the FNR remains 70.3%, while the FPR rises to 31.9%. The Tier 2 coverage gap is most pronounced on artifact cleanup (92.9% FNR), where agents naturally fall back to editing state files when the expected CLI is unavailable. These results highlight a coverage boundary worth examining: auto mode assumes dangerous actions transit the shell, but agents routinely achieve equivalent effects through file edits that the classifier does not evaluate.

Zimo Ji, Zongjie Li, Pingchuan Ma +3

Hong Kong University of Science and Technology · ETH Zurich

AI Agent Security Tool-Use Authorization Tool Access Control for LLM Agents AI Coding Agents AI Agent Security Benchmarks

Apr 4, 2026

AgenticFlict: A Large-Scale Dataset of Merge Conflicts in AI Coding Agent Pull Requests on GitHub

Software Engineering 3.0 marks a paradigm shift in software development, in which AI coding agents are no longer just assistive tools but active contributors. While prior empirical studies have examined productivity gains and acceptance patterns in AI-assisted development, the challenges associated with integrating agent-generated contributions remain less understood. In particular, merge conflicts, a fundamental aspect of collaborative software development, remain underexplored in this context. In this paper, we present AgenticFlict, a large-scale dataset of textual merge conflicts in AI coding agent pull requests (Agentic PRs). The dataset comprises 142K+ Agentic PRs collected from 59K+ repositories, of which 107K+ are successfully processed through deterministic merge simulation. Our pipeline identifies 29K+ PRs exhibiting merge conflicts, yielding a conflict rate of 27.67%, and extracts 336K+ fine-grained conflict regions across these instances. Our preliminary exploratory analysis indicates that merge conflicts are both frequent and often substantial in AI-generated contributions, with noticeable variation across agents, emphasizing the need to better understand and manage integration challenges in AI-assisted software development. The dataset, code and supplementary materials are available in zenodo: https://doi.org/10.5281/zenodo.19396916.

Daniel Ogenrwot, John Businge

University of Nevada Las Vegas Las Vegas, USA

AI Coding Agents AI-Assisted Software Development

Apr 2, 2026

REAP: Automatic Curation of Coding Agent Benchmarks from Interactive Production Usage

Production deployment of AI coding agents requires fast, reproducible evaluation signals. Existing industrial practices trade off speed and fidelity: online A/B testing takes weeks and risks user experience, shadow deployment yields signals that are not reproducible across runs, and public benchmarks diverge from production workloads in language distribution, prompt style, and codebase structure. This paper presents REAP (Relevance and Execution-Audited Pipeline), an automated curation pipeline that constructs production-derived benchmarks from real developer-agent sessions without manual labeling. Such curation, while in-distribution to production usage, runs into several challenges. Untestable prompts, misaligned tests, and test flakiness all compromise evaluation reliability. While tasks can be manually audited to ensure only high-quality tasks remain in the benchmark, this approach is infeasible in the monorepo setting: the build infrastructure state is often ephemeral in large monorepos and requires the benchmark to be continuously re-curated against the current codebase. As manual verification cannot be sustained at this cadence, REAP adds an automated verification layer using LLM-based task classification, agentic test-relevance validation, and multi-run stability checks to ensure the executable benchmark yields trustworthy signals. We use REAP to curate Harvest, a benchmark where each task feeds the coding agent a real developer prompt and verifies the resulting code change against fail-to-pass tests retrieved from production. Harvest's distribution spans more than four programming languages with a majority of tasks drawn from Hack. Model and harness evaluations reveal that solve rates range from 42.9% to 58.2% across five frontier models, surfacing capability differences that inform concrete deployment decisions.

Smriti Jha, Matteo Paltenghi, Chandra Maddila +3

Meta, USA

Coding Agent Benchmarks Benchmark Construction AI Agent Benchmarks AI Coding Agents Automated Evaluation

Apr 1, 2026

Agent psychometrics: Task-level performance prediction in agentic coding benchmarks

As the focus in LLM-based coding shifts from static single-step code generation to multi-step agentic interaction with tools and environments, understanding which tasks will challenge agents and why becomes increasingly difficult. This is compounded by current practice: agent performance is typically measured by aggregate pass rates on benchmarks, but single-number metrics obscure the diversity of tasks within a benchmark. We present a framework for predicting success or failure on individual tasks tailored to the agentic coding regime. Our approach augments Item Response Theory (IRT) with rich features extracted from tasks, including issue statements, repository contexts, solutions, and test cases, and introduces a novel decomposition of agent ability into LLM and scaffold ability components. This parameterization enables us to aggregate evaluation data across heterogeneous leaderboards and accurately predict task-level performance for unseen benchmarks, as well as unseen LLM-scaffold combinations. Our methods have practical utility for benchmark designers, who can better calibrate the difficulty of their new tasks without running computationally expensive agent evaluations.

Chris Ge, Daria Kryvosheieva, Daniel Fried +2

MIT · Fulcrum · Carnegie Mellon University

LLM Agent Evaluation AI Agent Evaluation AI Coding Agents Agent Benchmarks Item Response Theory

Mar 31, 2026

View-oriented Conversation Compiler for Agent Trace Analysis

We observe that an agent trace is a structured document. A coding agent session contains user turns, assistant text, chain of thought blocks, tool calls, tool results, subagent invocations, compaction boundaries, and harness injected directives, and may exceed ten thousand JSONL lines. Effective trace analysis requires a lossless record, a session level overview, and content retrieval with conversational roles. To this end, we propose VCC, namely View oriented Conversation Compiler, which lexes, parses, and lowers a raw JSONL log into three views based on one intermediate representation. The full view provides the lossless transcript and defines the line number coordinate system. The UI view reconstructs the interaction perceived by the user. The adaptive view selects relevant trace content and preserves role annotations and line range pointers. Line numbers are assigned before view lowering, ensuring that all pointers can be resolved to the full view. We evaluate VCC in a context engineering experiment on AppWorld by varying the input format of the reflector. Experiments with three model configurations and two test splits show that VCC improves aggregate task goal pass rates by 1.1 to 4.2 points, reduces reflector token consumption by one half to two thirds, and generates smaller memory files. These results demonstrate the effectiveness of trace format as an important component of context engineering infrastructure.

Lvmin Zhang, Maneesh Agrawala

Stanford University

LLM Context Management Reasoning Trace Analysis AI Coding Agents

Mar 31, 2026

IACDM: Interactive Adversarial Convergence Development Methodology -- A Structured Framework for AI-Assisted Software Development

Adoption of AI-assisted development in 2025 exposed a tool-agnostic failure pattern: experienced developers using frontier models were measurably slower while believing they were faster, and 10.3% of applications in one production showcase leaked data through misconfigured access. These failures share a structural cause, the verification gap: absent external tool use, no language model can determine whether what it generated is correct. The tool is irrelevant; the process is determinative. We present IACDM (Interactive Adversarial Convergence Development Methodology), an 8-phase framework in which verification agents external to the generator operate at discrete gates, and the AI alternates between building artifacts and attacking them through specialized critique lenses. What distinguishes it from the review and red-teaming traditions it borrows from is that the gate is enforced by a state machine outside the model: the agent may request advancement, not grant it. One component has now been tested. A pre-registered experiment over twelve projects under a single frozen instrument asked whether the nineteen lenses are non-redundant, against a prior expectation from replications of Perspective-Based Reading, which found reading perspectives not to differ. Every lens found at least one defect no other lens found, under four independent judgements of what counts as the same defect. Two further results were unplanned: the taxonomy withstands testing while the criterion deciding when to apply it does not, and the method creates commitments nothing ever re-confronts with the built system. The method was otherwise developed at one industrial R&D institute. That evidence establishes applicability, not effectiveness: no project ran without the method, and the proponents are the evaluators. Whether IACDM outperforms direct prompting remains open, and this paper is written to make it refutable.

Jasmine Moreira

AI-Assisted Software Development AI Coding Agents Automated Code Review

Mar 29, 2026

Needle in the Repo: A Benchmark for Maintainability in AI-Generated Repository Edits

AI coding agents can now complete complex programming tasks, but existing evaluations largely emphasize behavioral correctness and often overlook maintainability risks such as weak modularity or testability. We present Needle in the Repo (NITR), a diagnostic probe-and-oracle framework for evaluating whether behaviorally correct repository edits preserve maintainable structure. NITR distills recurring software engineering wisdom into controlled probes embedded in small, realistic multi-file codebases, each designed so that success depends primarily on one targeted maintainability dimension. Each probe is paired with a hidden evaluation harness that combines functional tests for required behavior with structural oracles that encode the targeted maintainability constraint and return interpretable diagnoses. Using NITR, we evaluate 23 coding configurations across GPT, Claude, Gemini, and Qwen families in both direct-inference and agent-based settings. Current AI coding systems remain far from robust: on average, configurations solve only 36.2% of cases, the best reaches 57.1%, and performance drops from 53.5% on micro cases to 20.6% on multi-step cases. The hardest pressures are architectural rather than local edits, especially dependency control (4.3%) and responsibility decomposition (15.2%). Moreover, 64/483 outcomes (13.3%) pass all functional tests yet fail the structural oracle. Under our harness, agent-mode configurations improve average performance from 28.2% to 45.0%, but do not eliminate these architectural failures. These results show that progress in code generation is not yet progress in maintainable code evolution, and that NITR exposes a critical failure surface missed by conventional evaluation.

Haichao Zhu, Qian Zhang, Jiyuan Wang +2

Independent Researcher · UC Riverside · Tulane University

Software Engineering AI Coding Agents Code Generation Benchmarks Repository-Level Code Generation

Mar 29, 2026

KAT-Coder-V2 Technical Report

We present KAT-Coder-V2, an agentic coding model developed by the KwaiKAT team at Kuaishou. KAT-Coder-V2 adopts a "Specialize-then-Unify" paradigm that decomposes agentic coding into five expert domains - SWE, WebCoding, Terminal, WebSearch, and General - each undergoing independent supervised fine-tuning and reinforcement learning, before being consolidated into a single model via on-policy distillation. We develop KwaiEnv, a modular infrastructure sustaining tens of thousands of concurrent sandbox instances, and scale RL training along task complexity, intent alignment, and scaffold generalization. We further propose MCLA for stabilizing MoE RL training and Tree Training for eliminating redundant computation over tree-structured trajectories with up to 6.2x speedup. KAT-Coder-V2 achieves 79.6% on SWE-bench Verified (vs. Claude Opus 4.6 at 80.8%), 88.7 on PinchBench (surpassing GLM-5 and MiniMax M2.7), ranks first across all three frontend aesthetics scenarios, and maintains strong generalist scores on Terminal-Bench Hard (46.8) and tau^2-Bench (93.9). Our model is publicly available at https://streamlake.com/product/kat-coder.

Fengxiang Li, Han Zhang, Haoyang Huang +43

AI Coding Agents Agentic Code Generation LLM Agent Training RL for Code Generation

Mar 29, 2026

PRBench: End-to-end Paper Reproduction in Physics Research

AI agents powered by large language models exhibit strong reasoning and problem-solving capabilities, enabling them to assist scientific research tasks such as formula derivation and code generation. However, whether these agents can reliably perform end-to-end reproduction from real scientific papers remains an open question. We introduce PRBench, a benchmark of 30 expert-curated tasks spanning 11 subfields of physics. Each task requires an agent to comprehend the methodology of a published paper, implement the corresponding algorithms from scratch, and produce quantitative results matching the original publication. Agents are provided only with the task instruction and paper content, and operate in a sandboxed execution environment. All tasks are contributed by domain experts from over 20 research groups at the School of Physics, Peking University, each grounded in a real published paper and validated through end-to-end reproduction with verified ground-truth results and detailed scoring rubrics. Using an agentified assessment pipeline, we evaluate a set of coding agents on PRBench and analyze their capabilities across key dimensions of scientific reasoning and execution. The best-performing agent, OpenAI Codex powered by GPT-5.3-Codex, achieves a mean overall score of 34%. All agents exhibit a zero end-to-end callback success rate, with particularly poor performance in data accuracy and code correctness. We further identify systematic failure modes, including errors in formula implementation, inability to debug numerical simulations, and fabrication of output data. Overall, PRBench provides a rigorous benchmark for evaluating progress toward autonomous scientific research.

Shi Qiu, Junyi Deng, Yiwei Deng +48

AI Agent Benchmarks AI Coding Agents Scientific Reproducibility Scientific Reasoning in Language Models

Mar 29, 2026

Safer Builders, Risky Maintainers: A Comparative Study of Breaking Changes in Human vs Agentic PRs

AI coding agents are increasingly integrated into modern software engineering workflows, actively collaborating with human developers to create pull requests (PRs) in open-source repositories. Although coding agents improve developer productivity, they often generate code with more bugs and security issues than human-authored code. While human-authored PRs often break backward compatibility, leading to breaking changes, the potential for agentic PRs to introduce breaking changes remains underexplored. The goal of this paper is to help developers and researchers evaluate the reliability of AI-generated PRs by examining the frequency and task contexts in which AI agents introduce breaking changes. We conduct a comparative analysis of 7,191 agent-generated PRs with 1402 human-authored PRs from Python repositories in the AIDev dataset. We develop a tool that analyzes code changes in commits corresponding to the agentic PRs and leverages an abstract syntax tree (AST) based analysis to detect potential breaking changes. Our findings show that AI agents introduce fewer breaking changes overall than humans (3.45% vs. 7.40%) in code generation tasks. However, agents exhibit substantially higher risk during maintenance tasks, with refactoring and chore changes introducing breaking changes at rates of 6.72% and 9.35%, respectively. We also identify a "Confidence Trap" where highly confident agentic PRs still introduce breaking changes, indicating the need for stricter review during maintenance oriented changes regardless of reported confidence score.

K M Ferdous, Dipayan Banik, Kowshik Chowdhury +1

Kennesaw State University Marietta, Georgia, USA · Quanta Technology Raleigh, North Carolina, USA

AI Coding Agents LLM Agent Reliability

Mar 28, 2026

A Multi-agent AI System for Deep Learning Model Migration from TensorFlow to JAX

The rapid development of AI-based products and their underlying models has led to constant innovation in deep learning frameworks. Google has been pioneering machine learning usage across dozens of products. Maintaining the multitude of model source codes in different ML frameworks and versions is a significant challenge. So far the maintenance and migration work was done largely manually by human experts. We describe an AI-based multi-agent system that we built to support automatic migration of TensorFlow-based deep learning models into JAX-based ones. We make three main contributions: First, we show how an AI planner that uses a mix of static analysis with AI instructions can create migration plans for very complex code components that are reliably followed by the combination of an orchestrator and coders, using AI-generated example-based playbooks. Second, we define quality metrics and AI-based judges that accelerate development when the code to evaluate has no tests and has to adhere to strict style and dependency requirements. Third, we demonstrate how the system accelerates code migrations in a large hyperscaler environment on commercial real-world use-cases. Our approach dramatically reduces the time (6.4x-8x speedup) for deep learning model migrations and creates a virtuous circle where effectively AI supports its own development workflow. We expect that the techniques and approaches described here can be generalized for other framework migrations and general code transformation tasks.

Stoyan Nikolov, Bernhard Konrad, Moritz Gronbach +5

Google Munich, Germany · Google Mountain View, California, USA · Google Sunnyvale, California, USA

AI Coding Agents Multi-Agent Coding

Mar 28, 2026

Codebase-Memory: Tree-Sitter-Based Knowledge Graphs for LLM Code Exploration via MCP

Large Language Model (LLM) coding agents typically explore codebases through repeated file-reading and grep-searching, consuming thousands of tokens per query without structural understanding. We present Codebase-Memory, an open-source system that constructs a persistent, Tree-Sitter-based knowledge graph via the Model Context Protocol (MCP), parsing 66 languages through a multi-phase pipeline with parallel worker pools, call-graph traversal, impact analysis, and community discovery. Evaluated across 31 real-world repositories, Codebase-Memory achieves 83% answer quality versus 92% for a file-exploration agent, at ten times fewer tokens and 2.1 times fewer tool calls. For graph-native queries such as hub detection and caller ranking, it matches or exceeds the explorer on 19 of 31 languages.

Martin Vogel, Falk Meyer-Eschenbach, Severin Kohler +2

Independent Researcher, Berlin, Germany · Institute of Medical Informatics, Charité – Universitätsmedizin Berlin, Berlin, Germany · Clinical Study Center, Berlin Institute of Health, Berlin, Germany +3

AI Coding Agents Graph-Based Agent Memory Knowledge Graphs

Mar 27, 2026

Ask or Assume? Uncertainty-Aware Clarification-Seeking in Coding Agents

As Large Language Model (LLM) agents are increasingly deployed in open-ended domains like software engineering, they frequently encounter underspecified instructions that lack crucial context. While human developers naturally resolve underspecification by asking clarifying questions, current agents are largely optimized for autonomous execution. In this work, we systematically evaluate the clarification-seeking abilities of LLM agents on an underspecified variant of SWE-bench Verified. We propose an uncertainty-aware multi-agent scaffold that decouples underspecification detection from code execution. Across both proprietary and open-weight frontier LLMs, our scaffold achieves a 69.40% task resolve rate, significantly outperforming a standard single-agent setup and closing the performance gap with agents operating on fully specified instructions. Furthermore, we find that the multi-agent system exhibits well-calibrated information-seeking behavior, conserving queries on simple tasks while proactively seeking information on more complex issues. These findings indicate that current models can be turned into proactive collaborators, where agents independently recognize when to ask questions to elicit missing information in real-world, underspecified tasks.

Nicholas Edwards, Sebastian Schuster

Faculty of Computer Science, University of Vienna, Vienna, Austria · UniVie Doctoral School Computer Science, University of Vienna, Vienna, Austria

Multi-Agent LLM Systems AI Coding Agents LLM Agents LLM Uncertainty Estimation

Mar 26, 2026

Confident and Wrong: Silent Semantic Failures in Coding Agents

As coding agents move into production workflows, teams need to know not only whether an agent completes a task, but whether its action can be trusted. We show that completion and trustworthiness diverge sharply and systematically. Across 1,750 trajectories on 50 SWE-bench Verified tasks, we compare four frontier models over repeated runs and separate submit rate from test-verified resolve rate. GPT-5 submits a patch on 100% of runs but resolves only 44%; Llama 4 submits on 99% but resolves 18%; and Gemini, despite submitting least often at 70%, resolves more tasks than GPT-5 (50% versus 44%). These gaps are not random: they concentrate in one dangerous failure mode we call silent semantic failure. Qualitatively, on a buggy task the agent submits a plausible-looking patch on all five runs, yet none pass, the same misinterpretation repeated rather than random error. Quantitatively, it dominates failure, covering 80% of Llama 4's failing runs and 68% of GPT-5's, and it is invisible: the outcomes are confidently and consistently wrong, so completion-based and consistency-based monitoring both look healthy exactly when the agent should not be trusted. Lightweight pre-edit prompts do not close the gap. A second probe isolates the instinct to act: given an already-fixed bug, where the right move is to abstain, most models still edit the correct code. This action bias, acting when no action is warranted, is exactly what completion metrics reward. The throughline is measurement: submit rate captures action, but trust requires validity. So evaluation must catch up: score agents by test-verified correctness over repeated runs, report its uncertainty, and reward those that know when not to act.

Aman Mehta

Snowflake AI Research

AI Agent Reliability AI Coding Agents LLM Agent Evaluation LLM Agent Reliability