PulseAugur
EN
LIVE 01:56:37

New research explores optimizing agent harnesses for improved AI performance · 9 sources tracked

Multiple research papers explore the optimization and evolution of agent harnesses, which are crucial components that dictate how language models interact with their environments. Studies introduce novel frameworks like GUI-HARVEST, ActiveSaddler, MILO, and DynaHarness, focusing on self-improvement, automated curriculum learning, and dynamic physical governance. These methods aim to enhance agent performance across various tasks, from coding and general agent tasks to robotics, by refining prompts, tool interfaces, and control logic based on execution feedback and evidence-driven evolution. Findings suggest that the effectiveness of agent systems is highly dependent on the interplay between the model and its harness, with tailored harness optimization leading to significant performance gains. AI

IMPACT Advances in harness optimization could significantly improve the reliability and performance of AI agents across diverse applications.

RANK_REASON Multiple papers published on arXiv detailing new research into agent harness optimization techniques.

Read on arXiv cs.MA (Multiagent) →

AI-generated summary · Google Gemini · from 30 sources. How we write summaries →

New research explores optimizing agent harnesses for improved AI performance · 9 sources tracked

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
Multiple papers published on arXiv detailing new research into agent harness optimization techniques.
Source corroboration
30 independent sources
Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics
paper, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
56 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.
Coverage growth since scoring
+10 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

Full methodology in our editorial standards.

COVERAGE [30]

  1. arXiv cs.CL TIER_1 English(EN) · Haolin Yang, Jipeng Zhang, Jian Xie, Shuaishuai Gong, Sirui Han, Yike Guo ·

    HarnessSQL: Harness-Native Training for SQL Agents in Realistic Database Environments

    arXiv:2610.12274v1 Announce Type: new Abstract: Text-to-SQL models are commonly trained to map questions directly to static queries, whereas real-world database agents operate through stateful, multi-turn interaction with live databases -- inspecting schemas, executing probe quer…

  2. arXiv cs.CL TIER_1 English(EN) · Jixuan Chen, Jiaxin Zhang, Qinyuan Ye, Yada Pruksachatkun, Haoxiang Zhang, Jingming Zhuo, Yifan Zhang, Yutong Dai, Juntao Tan, Xiangyu Peng, Silvio Savarese, Zeyuan Chen, Lianhui Qin, Chien-Sheng Wu ·

    CoTrace: Data Recipes for Training Terminal Agents with Harness-Model Co-Evolution

    arXiv:2610.10426v1 Announce Type: new Abstract: Terminal-agent capability depends jointly on model weights and the runtime harness that formats prompts, binds tools, and handles error recovery. Existing harness-model co-evolution approaches improve both components, yet often trea…

  3. Hugging Face Daily Papers TIER_1 English(EN) ·

    CoTrace: Data Recipes for Training Terminal Agents with Harness-Model Co-Evolution

    Terminal-agent capability depends jointly on model weights and the runtime harness that formats prompts, binds tools, and handles error recovery. Existing harness-model co-evolution approaches improve both components, yet often treat trajectories produced during harness search as…

  4. arXiv cs.AI TIER_1 Deutsch(DE) · Zekai Wang, Yingqiang Ge, Zekun Wang, Hai Wang, Yuhui Xu, Joshua Frandsen, Shancong Fu, Ashia C. Wilson, Chandan K. Reddy ·

    VERSE: Verified Self-Evolving Optimizer for Agent Harnesses

    arXiv:2610.02616v1 Announce Type: new Abstract: Harness evolution improves an LLM agent's prompts, tools, and workflow, while the optimizer's own tools and procedures often remain fixed. We study whether an optimizer can improve another agent more effectively by also improving ho…

  5. arXiv cs.AI TIER_1 English(EN) · Yixuan Li, Yiyun Zhou, Yao Long Teng, Fuchao Yang, Yanchen Deng, Zhiyi Lyu, Xuyu Dong, Feng Chen, Bo An ·

    Finding the Right Fit: Model-Harness Interactions across Agent Tasks

    arXiv:2610.00917v1 Announce Type: new Abstract: Choosing an agent system means choosing both a language model and the harness through which it acts. We ask whether a strong model, harness, or pairing stays strong when the setting changes. We evaluate 66 configurations: four confi…

  6. arXiv cs.AI TIER_1 English(EN) · Geyi Yang, Zikun Qu, Xiang Li, Zhiyong Wang, Min Zhang, Shipei Zeng, Zhongxiang Dai ·

    GUI-HARVEST: Self-Improving GUI Agents through Evidence-Driven Harness Evolution

    arXiv:2610.00948v1 Announce Type: cross Abstract: The executable harness surrounding a GUI model determines how observations are assembled, actions are executed, and verification, recovery, and termination are controlled. Compared with harness optimization for non-GUI agents, aut…

  7. arXiv cs.AI TIER_1 English(EN) · Sungho Park, Wonjoong Kim, Jue Zhang, Wook-Shin Han, Pengfei Gao, Chanyoung Park, Yongqiang Yao, Rao Fu, Elsie Nallipogu, Qingwei Lin, Victor R\"uhle ·

    ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization

    arXiv:2610.00906v1 Announce Type: new Abstract: Automated harness optimization can substantially improve LLM agents by iteratively updating their prompts, tool interfaces, and control logic from execution feedback. However, existing methods primarily optimize how the harness is u…

  8. Hugging Face Daily Papers TIER_1 English(EN) ·

    Dynamic Harness Search: Building Multi-Agent Systems Per-Query via Prediction

    Agent harnesses specify the roles, instructions, tools, and communication structure used to solve a task, and the right harness depends on the query. Because the value of each design choice is observable only through execution, tailoring a harness to each query has required eithe…

  9. arXiv cs.AI TIER_1 English(EN) · Haoyuan Deng, Jiebin Liu, Tengxiao Zhang, Langning Yan, Hongye Cao, Ziwei Wang ·

    DynaHarness: A Dynamic Physical Harness for Self-Evolving Robot Agents

    arXiv:2609.40306v1 Announce Type: cross Abstract: Pretrained robot policies provide useful action priors, but long-horizon manipulation still requires coordination between semantic reasoning and physical execution. Semantic reasoning operates at a coarser timescale than physical …

  10. arXiv cs.AI TIER_1 English(EN) · Zhening Li, Joshua Liu, Mateja Vukelic, Nicole Shen, Supriya Lall, Amitayush Thakur, Alex Zhang, Omar Khattab, Jonathan Light, Armando Solar-Lezama ·

    Harness as a Language: A Minimalist Agent Framework With Maximal Expressivity

    arXiv:2609.26891v2 Announce Type: replace Abstract: Modern language-model agents are built around the agent loop: the LLM is placed in an environment exposing a set of tools, and the LLM has full control over the workflow by alternating between tool calls and observing their outp…

  11. arXiv cs.AI TIER_1 English(EN) · Zhixuan Tan, Pengjie Gu, Zhao Li, Yihan Hu, Xu He, Dong Li, Jianye Hao ·

    EngramBench: A Capability-Grounded Benchmark for Skill-Evolution Harnesses

    arXiv:2609.39284v1 Announce Type: cross Abstract: While large language models have achieved remarkable success in isolated code generation, authentic software engineering requires sustained reasoning, complex state management, and continuous cross-domain abstraction. However, cur…

  12. arXiv cs.AI TIER_1 English(EN) · Prithwish Jana, Mononito Goswami, Hao Liu, Xinyu Li, Langlin Huang, Zhehui Huang, Zhishen Huang, Patrick Bl\"obaum, Anoop Deoras, Purak Jain, Nikos Kanakaris, Sahika Genc ·

    MILO: Automated Harness Discovery via Orchestrated Multi-Agent Evolution

    arXiv:2609.38349v1 Announce Type: cross Abstract: Modern agentic systems combine an AI model with a harness that controls execution and environmental interactions. Harness design strongly affects long-horizon performance, yet its combinatorial search space demands substantial hum…

  13. arXiv cs.AI TIER_1 English(EN) · Jingbo Yang, Kwei-Herng Lai, Xiaowen Wang, Yaar Harari, Evgeniy Gabrilovich, Shiyu Chang ·

    Learning from Research: Toward Lifelong Agent Harness Evolution

    arXiv:2609.40169v1 Announce Type: new Abstract: Language agents are expected to solve increasingly complex tasks, creating a growing need for continual improvement. One promising approach is to evolve the agent harness, the software that governs tool use, memory management, and t…

  14. arXiv cs.AI TIER_1 English(EN) · Chen Lu, Ke Xue, Siyuan Xu, Mingxuan Yuan, Chao Qian ·

    Self-Evolving Algorithm-Design Agents: Escaping In-Context Evolutionary Stagnation via Population-Curated Policy Optimization

    arXiv:2609.38757v1 Announce Type: new Abstract: Large language models are increasingly participating in complex real-world tasks in the form of algorithm-design agents, designing and refining algorithms. Many successful algorithm-design agents adopt pure in-context evolutionary f…

  15. arXiv cs.AI TIER_1 English(EN) · Qiankai Xu ·

    Self-Evolving Harness on Multiple Tasks with the Agent as Its Own Optimizer

    arXiv:2609.38372v1 Announce Type: new Abstract: A harness is the code around a language-model agent that organizes prompts, calls tools, manages context, and controls execution. As models grow stronger, recent work has begun to let agents improve their own harnesses, a line of wo…

  16. arXiv cs.AI TIER_1 English(EN) · Minki Kang, Ryo Hachiuma, Shaokun Zhang, Subhashree Radhakrishnan, Yonggan Fu, Jindong Jiang, Mingjie Liu, Ehsan Hosseini-Asl, Yi Dong, Yu-Chiang Frank Wang, Byung-Kwan Lee ·

    Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents

    arXiv:2609.39982v1 Announce Type: cross Abstract: Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable execution. A poor command (e.g., wrong package install) can change the environment in ways that hin…

  17. arXiv cs.AI TIER_1 English(EN) · Zhijie Wei, Ferris Tan, Jinghui Wang ·

    Scale and Selection: What Makes Automatic Harness Evolution Work for Visual-Interface Robot Agents

    arXiv:2609.39304v1 Announce Type: cross Abstract: When an off-the-shelf coding agent is used directly as a robot policy, observing a browser-based 3D interface through screenshots and acting by posing a virtual target gripper through a few tools, the agent's harness, its prompts,…

  18. arXiv cs.MA (Multiagent) TIER_1 English(EN) · Victor Rühle ·

    ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization

    Automated harness optimization can substantially improve LLM agents by iteratively updating their prompts, tool interfaces, and control logic from execution feedback. However, existing methods primarily optimize how the harness is updated while largely fixing which training scena…

  19. Hugging Face Daily Papers TIER_1 English(EN) ·

    ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization

    Automated harness optimization can substantially improve LLM agents by iteratively updating their prompts, tool interfaces, and control logic from execution feedback. However, existing methods primarily optimize how the harness is updated while largely fixing which training scena…

  20. Hugging Face Daily Papers TIER_1 English(EN) ·

    GUI-HARVEST: Self-Improving GUI Agents through Evidence-Driven Harness Evolution

    The executable harness surrounding a GUI model determines how observations are assembled, actions are executed, and verification, recovery, and termination are controlled. Compared with harness optimization for non-GUI agents, automatically optimizing this harness poses three cou…

  21. Hugging Face Daily Papers TIER_1 English(EN) ·

    Learning from Research: Toward Lifelong Agent Harness Evolution

    Language agents are expected to solve increasingly complex tasks, creating a growing need for continual improvement. One promising approach is to evolve the agent harness, the software that governs tool use, memory management, and task execution, while keeping the underlying lang…

  22. arXiv cs.MA (Multiagent) TIER_1 English(EN) · Byung-Kwan Lee ·

    Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents

    Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable execution. A poor command (e.g., wrong package install) can change the environment in ways that hinder subsequent progress, even when the model could…

  23. arXiv cs.AI TIER_1 English(EN) · Haoyu Dong, Yuhang Zhou, Zihao Lin, Yifan Wu, Bo Peng, Mingyi Wang, Xiangjun Fan, Lizhu Zhang, Zhuokai Zhao ·

    Mixture of Self-Improving Branches For Agent Harness Optimization

    arXiv:2609.37834v1 Announce Type: new Abstract: Harness optimization provides a practical setting for recursive self-improvement (RSI), where agent-generated modifications inform subsequent changes through execution feedback. Recent work such as Meta-Harness implements this proce…

  24. arXiv cs.NE (Neural & Evolutionary) TIER_1 English(EN) · Chao Qian ·

    Self-Evolving Algorithm-Design Agents: Escaping In-Context Evolutionary Stagnation via Population-Curated Policy Optimization

    Large language models are increasingly participating in complex real-world tasks in the form of algorithm-design agents, designing and refining algorithms. Many successful algorithm-design agents adopt pure in-context evolutionary frameworks, but they may quickly plateau in domai…

  25. Hugging Face Daily Papers TIER_1 English(EN) ·

    Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents

    Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable execution. A poor command (e.g., wrong package install) can change the environment in ways that hinder subsequent progress, even when the model could…

  26. Hugging Face Daily Papers TIER_1 English(EN) ·

    MILO: Automated Harness Discovery via Orchestrated Multi-Agent Evolution

    Modern agentic systems combine an AI model with a harness that controls execution and environmental interactions. Harness design strongly affects long-horizon performance, yet its combinatorial search space demands substantial human effort that must be repeated as models change. …

  27. Hugging Face Daily Papers TIER_1 English(EN) ·

    PluginRSI: Recursive Improvement of Agent Harnesses with Reusable Plugins

    The harness surrounding a language model is a central determinant of agent performance. Recent methods optimize harnesses by searching over complete programs, where individual mechanisms are difficult to isolate and reuse. We introduce PluginRSI, which represents a harness as a c…

  28. dev.to — MCP tag TIER_1 English(EN) · Sravya Dangeti ·

    Robot-Use Agents: Why the Harness Matters More Than the Model

    <h2> TL;DR </h2> <p>General-purpose LLM agents can now control robots with little or no robot-specific training. The hard engineering work is moving from the model to the harness around it: the tools it can call, the checks that run before it acts, and the evals that tell you whe…

  29. dev.to — LLM tag TIER_1 English(EN) · Takashi Abe ·

    "Is That a Harness Too?" Making Sense of Harness Engineering for AI Agents

    <h2> When sales says "everything is a harness," they're not wrong. It's just too broad </h2> <p>Lately I keep hearing the same lines from sales folks around me: "Oh, that's harness engineering," or "Are you really thinking about harness engineering?" What exactly do they mean by …

  30. Mastodon — mastodon.social TIER_1 Русский(RU) · [email protected] ·

    🔍 HarnessRouter — a unified FOSS interface for any harness agents (Codex, Claude Code, Hermes, DeepSeek Harness, System One by Jev, and 9 more) turning them into

    🔍 HarnessRouter — единый FOSS интерфейс для любых харнессов агентов (Codex, Claude Code, Hermes, DeepSeek Harness, System One от Jev и ещё 9) превращающий их в plug-and-play агентные бэкенды. Запускает любой харнесс под единым интерфейсом: продукт передаёт задачи и получает резул…