Pith. sign in

REVIEW 30 cited by

Trae Agent: An LLM-based Agent for Software Engineering with Test-time Scaling

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2507.23370 v1 pith:HDS4BPRI submitted 2025-07-31 cs.SE cs.AI

Trae Agent: An LLM-based Agent for Software Engineering with Test-time Scaling

classification cs.SE cs.AI
keywords agenttraeensemblesoftwareengineeringissuelargereasoning
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Software issue resolution is a critical challenge in software engineering and has garnered increasing attention in recent years. With the rapid advancement of large language models (LLMs), substantial progress has been made in addressing real-world software engineering tasks. Recent studies have introduced ensemble reasoning techniques to enhance the performance of LLM-based issue resolution. However, existing prompting-based methods still face limitations in effectively exploring large ensemble spaces and lack the capacity for repository-level understanding, both of which constrain their overall effectiveness. In this paper, we propose Trae Agent, the first agent-based ensemble reasoning approach for repository-level issue resolution. Trae Agent formulates our goal as an optimal solution search problem and addresses two key challenges, i.e., large ensemble spaces and repository-level understanding, through modular agents for generation, pruning, and selection. We conduct extensive experiments using three leading LLMs on the widely-adopted SWE-bench benchmark, comparing Trae Agent against four state-of-the-art ensemble reasoning techniques. Experimental results demonstrate that Trae Agent consistently achieves superior performance, with an average improvement of 10.22% over all baselines in terms of Pass@1. Trae Agent has achieved first place on the SWE-bench Verified leaderboard, with a notable Pass@1 score of 75.20%. We are pleased to release Trae Agent as an open-source project to support the research community, with all resources available at https://github.com/bytedance/trae-agent.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 30 Pith papers

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

  1. When Do Intrinsic Rewards Work for Code Reasoning? A Comprehensive Study

    cs.AI 2026-06 unverdicted novelty 7.0

    Empirical evaluation on LiveCodeBench shows certainty-based RLIF yields early gains followed by output shortening and reasoning collapse, providing no advantage for RLVR initialization on code tasks.

  2. SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering

    cs.SE 2026-05 unverdicted novelty 7.0

    SaaSBench introduces a heterogeneous benchmark for enterprise SaaS engineering and shows that state-of-the-art coding agents fail over 95% of the time before reaching deep business logic due to setup and integration problems.

  3. SWE-WebDevBench: Evaluating Coding Agent Application Platforms as Virtual Software Agencies

    cs.MA 2026-05 conditional novelty 7.0

    SWE-WebDevBench finds that AI app builders commonly fail at translating business needs into complete, secure, production-ready software due to specification bottlenecks, frontend-backend decoupling, low engineering qu...

  4. Evaluating LLM Agents on Automated Software Analysis Tasks

    cs.SE 2026-04 unverdicted novelty 7.0

    A custom LLM agent achieves 94% manually verified success on a new benchmark of 35 software analysis setups, outperforming baselines at 77%, but struggles with stage mixing, error localization, and overestimating its ...

  5. Can You Trust the Vectors in Your Vector Database? Black-Hole Attack from Embedding Space Defects

    cs.CR 2026-04 unverdicted novelty 7.0

    Injecting a few malicious vectors near the centroid exploits centrality-driven hubness in high-dimensional embeddings, causing them to dominate top-k retrievals in up to 99.85% of cases.

  6. From Component Manipulation to System Compromise: Understanding and Detecting Malicious MCP Servers

    cs.CR 2026-04 unverdicted novelty 7.0

    Presents a component-centric PoC dataset of malicious MCP servers and a two-stage behavioral deviation detector Connor achieving 94.6% F1-score.

  7. Automating Database-Native Function Code Synthesis with LLMs

    cs.DB 2026-04 conditional novelty 7.0

    DBCooker automates synthesis of database native functions via LLM-guided characterization, coding plans, hybrid filling, and progressive validation, delivering 34.55% higher accuracy than baselines on SQLite, PostgreS...

  8. Investigating Test Overfitting on SWE-bench

    cs.SE 2025-11 unverdicted novelty 7.0

    The first empirical study of test overfitting shows that auto-generated tests from issues can lead to code that passes observed tests but misses important cases or breaks functionality in SWE-bench issue resolution.

  9. Writing Bug Reports for Software Repair Agents: What Information Matters Most?

    cs.SE 2026-07 conditional novelty 6.5

    Localization cues and suggested fixes most strongly raise software-repair agents’ success on SWE-bench Verified, while classic human-oriented details like reproduction steps matter less when operational cues remain.

  10. SpecFirst: Behavioral Specification Elicitation as a First-Class Step in Agent-Based Program Synthesis from Scratch

    cs.SE 2026-07 conditional novelty 6.0

    A two-stage SpecFirst pipeline that elicits a structured behavioral SPEC.md before code synthesis significantly improves ProgramBench pass rates and binary exploration coverage over single-loop agents.

  11. CoACT: Action-Preserving Observation Compression for Coding Agents

    cs.SE 2026-07 conditional novelty 6.0

    CoACT trains a lightweight compressor via next-action preservation so coding agents use ~33% fewer tokens on SWE-bench Verified with little or no loss of task success.

  12. Beyond Textual Repository Exploration: Dual-Modal Structural Reasoning for Agentic Issue Resolution

    cs.SE 2026-07 unverdicted novelty 6.0

    DUALVIEW is a dual-modal framework using Module Coupling, Function Call, Class Hierarchy, and Program Dependence graphs to enable persistent structural reasoning for agentic issue resolution, reporting gains on SWE-be...

  13. LLVM-Bench: Benchmarking and Advancing Large Language Models for LLVM Compiler Issue Resolution

    cs.SE 2026-07 unverdicted novelty 6.0

    LLVM-Bench supplies 423 validated LLVM issues and LLVM-Gym automates evaluation, showing LLMs are limited but an ensemble reaches 21.99% resolution.

  14. CLIP: Lightweight Cosine-Law-Based Inverted-List Pruning for IVF-Based Vector Search

    cs.DB 2026-06 unverdicted novelty 6.0

    CLIP proposes a cosine-law-based pruning method for IVF vector search enabling O(1) cluster and log-time vector pruning with guarantees, plus variants for hierarchical and dynamic settings, showing up to 78% pruning a...

  15. A11YRepair: Bridging Web Accessibility Barriers via Knowledge-Enhanced Divide-and-Conquer Repair

    cs.SE 2026-06 unverdicted novelty 6.0

    A11YRepair uses divide-and-conquer clustering and WCAG knowledge to repair multi-fault accessibility violations in web projects, showing better results than baselines on a new 60-project benchmark with some patches me...

  16. From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flaws

    cs.SE 2026-06 unverdicted novelty 6.0

    HarnessFix is a trace-guided framework that diagnoses failures in LLM agent harnesses via a new intermediate representation and applies scoped repairs, yielding 15-50% gains on four benchmarks.

  17. Improving LLM Code Generation via Requirement-Aware Curriculum Reinforcement Learning

    cs.SE 2026-05 unverdicted novelty 6.0

    REC RL improves LLM code generation by automatically assessing and optimizing requirement difficulty with adaptive curriculum sampling, yielding 1.23-5.62% Pass@1 gains over baselines.

  18. When LLMs Lag Behind: Knowledge Conflicts from Evolving APIs in Code Generation

    cs.SE 2026-04 unverdicted novelty 6.0

    LLMs produce executable code only 42.55% of the time under API evolution without full documentation, improving to 66.36% with structured docs and by 11% more with reasoning strategies, yet outdated patterns persist.

  19. REAgent: Requirement-Driven LLM Agents for Software Issue Resolution

    cs.SE 2026-04 unverdicted novelty 6.0

    REAgent improves LLM patch generation for software issues by 17.4% on average through automated construction, quality checking, and iterative refinement of structured issue-oriented requirements.

  20. On the Role of Fault Localization Context for LLM-Based Program Repair

    cs.SE 2026-04 unverdicted novelty 6.0

    More fault localization context does not consistently improve LLM-based program repair; file-level context gives 15-17x gains, optimal around 6-10 files, while line-level context often degrades performance from noise.

  21. Can You Trust the Vectors in Your Vector Database? Black-Hole Attack from Embedding Space Defects

    cs.CR 2026-04 unverdicted novelty 6.0

    Injecting a few vectors near the embedding-space centroid can make them appear in top-k results for up to 94.4% of queries via centrality-driven hubness.

  22. Beyond Fixed Tests: Repository-Level Issue Resolution as Coevolution of Code and Behavioral Constraints

    cs.SE 2026-04 unverdicted novelty 6.0

    Agent-CoEvo is a multi-agent LLM framework that coevolves code patches and test patches to resolve repository-level issues, outperforming fixed-test baselines on SWE-bench Lite and SWT-bench Lite.

  23. Combating Data Laundering in LLM Training

    cs.CR 2026-04 conditional novelty 6.0

    Data laundering collapses original-query memorization detectors; SDR recovers useful detection signals by synthesizing training-like rewrites of proprietary data via a goal-details search.

  24. Can Old Tests Do New Tricks for Resolving SWE Issues?

    cs.SE 2025-10 conditional novelty 6.0

    TestPrune minimizes regression test suites to improve bug reproduction and patch validation in LLM-based agentic repair pipelines, delivering 6-13% relative gains on SWE-Bench benchmarks at low API cost.

  25. From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flaws

    cs.SE 2026-06 unverdicted novelty 5.0

    HarnessFix diagnoses harness flaws from agent traces via HTIR, maps them to repair operators, and improves benchmark performance by 6.3-18.4% over baselines.

  26. CP-Agent: A Calibrated Risk-Controlled Agent for Feedback-Driven Competitive Programming

    cs.CL 2026-05 unverdicted novelty 5.0

    CP-Agent improves LLM competitive programming performance via calibrated feedback mechanisms that target false-admission risk, evidence against bad programs, and success hazard.

  27. "Refactoring Runaway": Understanding and Mitigating Tangled Refactorings in Coding Agents for Issue Resolution

    cs.SE 2026-05 unverdicted novelty 5.0

    Empirical study finds coding agents produce fewer and less intense tangled refactorings than humans on Multi-SWE-bench; a refactoring-aware refinement improves compilability from 19.34% to 38.33% and resolves 2.79% mo...

  28. KISS Sorcar: A Stupidly-Simple General-Purpose and Software Engineering AI Assistant

    cs.SE 2026-04 unverdicted novelty 5.0

    KISS Sorcar introduces a simple layered agent framework and VS Code IDE that reaches 62.2% pass rate on Terminal Bench 2.0 by combining ReAct execution, summarization-based continuation, parallel tools, persistent his...

  29. KISS Sorcar: A Stupidly-Simple General-Purpose and Software Engineering AI Assistant

    cs.SE 2026-04 unverdicted novelty 5.0

    The paper introduces KISS Sorcar, a simple open-source AI agent framework with a five-layer hierarchy and git worktree isolation to address context limits, error propagation, and reviewability in software engineering tasks.

  30. Rethinking the Value of Agent-Generated Tests for LLM-Based Software Engineering Agents

    cs.SE 2026-02 unverdicted novelty 5.0

    Agent-generated tests mainly act as observational feedback channels and do not meaningfully improve issue resolution success in current LLM software engineering agents.