architecture · first seen 5 Oct 2026
self-evolving agents
Also: agent self-evolution, harness self-evolution, retrospective adaptation, coding agent self-evolution
LLM-based agents that autonomously modify their own runtime instructions or tools to adapt to changing conditions.
Self-evolving agents are LLM-based agents that improve from their own experience by rewriting persistent state outside the model weights, such as skill memories, prompts, tool logic or the runtime harness, and sometimes by generating their own training tasks. Recent papers concentrate on making that loop safe and controllable. One paper confines self-modification to the harness so each change in a credit pipeline is a reviewable diff, and another gates memory updates with verifier-based falsification. A security paper shows how distilled skills can be poisoned. Other work applies the idea to medical reasoning, VLM skills and multi-user memory. A few GitHub frameworks such as EvoHarness-RL and EvoSprig package it for developers.
If agents keep adapting at deployment time without retraining, deciding which changes get admitted, audited and protected from poisoning becomes a central engineering and governance problem. Harness-level evolution could make continual agent improvement practical and reviewable in regulated settings.
Analyst note (emerging), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 18 on 16 Nov 2026 · 16 mentions in 28 days · 16 independent groups · code → papers · first seen 6 weeks ago · acceleration +1.45
Earliest sources
- EvoHarnessRL/EvoHarness-RLCode · 5 Oct 2026 · EvoHarnessRL · 32 stars
- MedZERO: Self-Evolving Agents for Open-Ended Medical Reasoning Through Controlled Knowledge AccumulationPapers · 6 Oct 2026 · Xilin Dang, Weilin Ruan, Xue Yang, Jinghao Wang, Xiaowei Hu, Jinpeng Li, Pheng-A…
- Self-Evolve With a Reference:Anchored Training of Tool-Integrated AgentsPapers · 7 Oct 2026 · Wenjie Liao, Liangjie Zhao, Zehong Cao
- The Harness as the Only Mutable Surface: Compliance-Bounded Self-Evolution of LLM Agents in Credit Pipelines, with a Measured Admission GatePapers · 7 Oct 2026 · Ravil Akhtyamov
- AgentEvolver: System-Wide Self-Evolution Through Task ExecutionPapers · 8 Oct 2026 · Wentao Zhang, Fuchao Yang, Yilei Zhao, Xinrun Wang, Bo An
- EvoBelief: Falsifiable World Models for Self-Improving LLM AgentsPapers · 8 Oct 2026 · Janet Wang, Jiang Guo, Yi Xiang, Jiaqing Yuan, Mingwen Dong, Hideo Kobayashi, Sh…
Latest sources
- MetaAct: Self-Evolving Executable Action Constraints for Embodied Language AgentsPapers · 16 Nov 2026 · Jiaxing Zhao, Hanwen Wang, Di Liu, Shaoteng Liu, Jing-Ran Su, Yanheng Li, Qian Z…
- Prior-as-Skill: Remember Outside and Read Inside with On-Policy Self-Distillation for Self-Evolving VLM AgentsPapers · 11 Nov 2026 · Yan Lin, Jingyu Sun, Yifei Qian, Zhengtao Yao, Yuxia Wang, Daizong Liu, Chun Pon…
- Verify by Doing: Co-Evolving Agent Execution and Evaluation without External FeedbackPapers · 11 Nov 2026 · Yitong Lin, Jianda Chen, Yihao Hu, Pan Wang, Hang Wang, Zhihao Wen, Xiaoxue Ren,…
- When Should Agents Remember? Falsification-Gated Self-Evolution for LLM AgentsPapers · 10 Nov 2026 · Runxuan Tang, Haoyu Gao, Yuyan Ding, Junyi Yao, Zihao Zheng, Zecheng Sheng, Liwe…
- A Cross-Scenario Survey and Analysis of Self-Evolving Multimodal AgentsPapers · 10 Nov 2026 · Jingjing Jiang
- When Experience Becomes a Backdoor: Poisoning Self-Evolving LLM Agents via Failure-Aware Task OptimizationPapers · 7 Nov 2026 · Minhua Lin, Zongyu Wu, Hanqing Lu, Xianfeng Tang, Xiang Zhang, Qi He, Suhang Wan…
- Who Remembers What? Multi-Granularity Memory Extraction for Multi-User, Multi-Agent InteractionPapers · 7 Nov 2026 · Yunzhe Wang, Mesut Erhan Unal, Sheng Zhang, Shabnam Ghadar, Peng Tang
- From Reflection to Co-Evolution: Self-Evolving Agents with Locally Verified MemoryPapers · 7 Nov 2026 · Xinxiang Yin, Xinglang Zhang, Haodong Chen, Zhiqing Cui, Weidong Tang, Yuxuan Wa…
- MDSKILL: Learning to Internalize Skill Memory through Multi-View Distillation for Efficient Dynamic Agent Self-EvolutionPapers · 3 Nov 2026 · Dian Jin, Jinluan Yang, Kairong Han, Baohong Li, Zijing Hu, Yi He, Wenlin Zhong,…
- Artifact-to-Capability: Diagnosing Breaks in Capability Formation of Self-Evolving Agents in a Controlled EnvironmentPapers · 3 Nov 2026 · Ziqian Ban, Haoxiang Yang, Yanfei Zhang, Shixin Liang, Haoran Yu, Xu Lin, Shaobi…
- Toyhom/EvoSprigCode · 2 Nov 2026 · Toyhom · 103 stars
- More Rules, Longer Lists? Benchmarking Self-Evolving Agents on a Hidden-Rule CurriculumPapers · 29 Oct 2026 · Chenjie Qiu, Boyan Xu, Shaobin Shi, Bingfeng Chen, Ruichu Cai, Dongsheng Luo