method · first seen 4 Oct 2026
agent harness evolution
Also: EMHO, llm-harness, agent harness, agentic harness, harness evolution, LLM agent harness, coding agent harness, harness optimization
Iteratively optimizing the surrounding scaffolding of an LLM agent using execution feedback.
Agent harness evolution treats the code around a fixed LLM (context construction, state, tool dispatch, verification, memory) as the object to optimize, using iterative proposal-and-evaluation loops driven by execution feedback. The sources show a cluster of methods papers: MESH-Harness recombines modular harness components with bandit-guided search, a causal improvement graph replaces proposer-centric meta-harness search, Self-Improvement Memory carries harness edits across benchmarks and models, and REINS applies a self-evolving harness to real-time trajectory planning. A Berkeley paper with Ion Stoica studies harness-side cost optimizations. GitHub repos from Didi, Reflection and individual developers build customizable coding harnesses. The term 'harness' itself is also used loosely for agent runtimes in general.
If model weights stay fixed while the scaffolding is searched automatically, much of agent performance and cost tuning could move from manual prompt and tool engineering to reproducible optimization loops. The Berkeley result that harness changes mainly affect cost, not accuracy, is a useful check on how much these methods can deliver.
Analyst note (emerging), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 11 on 16 Nov 2026 · 57 mentions in 28 days · 57 independent groups · papers → code → discussion · first seen 6 weeks ago · acceleration -1.07
Earliest sources
- Causal Improvement Graph for Agentic Harness OptimizationPapers · 4 Oct 2026 · Junjie Zhang, Shunyu Liu, Haoyu Wang, Ting-En Lin, Yongbin Li, Dacheng Tao
- MESH-Harness: Self-Improving Agent Harnesses via Bandit-Guided Compositional EvolutionPapers · 4 Oct 2026 · Zhiwei Shang, Yu Huo, Mingrong Gong, An Yan, Zikun Qu, Junhao Dong, Bryan Kian H…
- AIDotNet/pi.netCode · 4 Oct 2026 · AIDotNet · 128 stars
- Flyx7Kazami/AcanvasCode · 4 Oct 2026 · Flyx7Kazami · 34 stars
- LucasXinYu/harness-rsiCode · 4 Oct 2026 · LucasXinYu · 25 stars
- MedicalHarness: A Controlled Evaluation of LLMs and Agent Harnesses on Medical TasksPapers · 5 Oct 2026 · Ziqing Wang, Lili Zhao, Kaize Ding
Latest sources
- Harness Leverage Across Model Size: A Ladder of Five Local Agent Harnesses from 0.75B to 4.7BPapers · 16 Nov 2026 · Hao Wang, Ting Huang
- Anatomy of an Agent Harness: Decisions, Invariants and Corner CasesPapers · 15 Nov 2026 · Resham Joshi
- Self-Improvement Memory Enables Continual Recursive Self-ImprovementPapers · 13 Nov 2026 · Xin Yu, Fei Xia, Jiaqi Pan, Youbao Tang, Chengzhi Liu, Adams Yu, Sicong Jiang, Y…
- EvoX: Rethinking the Agent Harness as a Runtime PolicyPapers · 11 Nov 2026 · Haoyang Zhang, Qumeng Sun, Xiao Zhang, Xiang Liu, Zhidong Fan, Junjie Wang
- System Optimizations for Cost-Efficient Agent HarnessesPapers · 9 Nov 2026 · Jongseok Park, Jiaxiang Yu, Eric Zhou, Tony Hong, Sunga Kim, Yusuf Afifi, Ian Wo…
- REINS: A Self-Evolving Agent Harness for Real-Time Trajectory PlanningPapers · 7 Nov 2026 · Zhihong Cui, Hengyu Liu, Haoran Tang, Shijun Liu, Amir Taherkordi, Tor Skeie
- FuzzyKuji/OpenProgramCode · 6 Nov 2026 · FuzzyKuji · 97 stars
- BOXBOX: Recursive Harness Self-Improvement for Long-Running Performance OptimizationPapers · 6 Nov 2026 · Yuhao Tan, Hanyun Jiang, Fangkai Yang, Yuan Yao, Taolue Chen, Bo Qiao, Yu Kang,…
- When Routine Operations Break Architectural Pattern Relationships in LLM Agent HarnessesPapers · 6 Nov 2026 · Boming Xia, Qinghua Lu, Dino Sejdinovic, Liming Zhu
- Flex-RSI: Benchmarking Recursive Self-Improvement of Frontier AgentsPapers · 6 Nov 2026 · Jingdong Zhang, Jionghao Wang, Zhengming Yu, Yancheng Cai, Wei Chow, Chris Choy,…
- Beneath the Surface: What We Overlook When We Talk About Agent EvolutionPapers · 5 Nov 2026 · Xiaobo Wang, Zimu Wang, Siyu Ye, Jiaqi Li, Min Tang, Jun Bai, Yuling Shi, Ruobin…
- EnvOpt: Automated Optimization for Agent EnvironmentsPapers · 5 Nov 2026 · Mingxuan Li, Huanzhi Mao, Qiuyang Mang, Hanchen Li, Joseph E. Gonzalez, Alex Dim…