mlsanityData to 7 Oct 2026

From Memory to Skills: Evidence-Grounded Co-Evolution Governance for Long-Horizon LLM Agents

arXiv 2607.16621 ↗First posted 18 Jul 2026cs.CLcited by 2

On the map: LLM agents and tool use → Agent skill learning and evolution

Abstract (arXiv)

Existing memory systems for long-horizon LLM agents often retrieve prior traces as passive context rather than converting them into executable capabilities. In this paper, we propose MSCE, a training-free Memory--Skill Co-Evolution framework that organizes agent experience into grounded step traces, reusable procedural policies, and declarative environmental cognition. MSCE crystallizes evidence-backed L2 policies with positive estimated gain into callable skills that retain evidence links, applicability boundaries, decision guidance, verification rules, and reliability estimates. It further introduces reflection-weighted value backfilling, which propagates sparse terminal feedback through dense local self-reflections to produce evidence-calibrated trace values for governing memory and skill evolution. Experiments on EvoAgentBench and LoCoMo demonstrate that MSCE significantly outperforms state-of-the-art skill-augmented and memory-driven agent baselines, exhibiting strong cross-domain transferability and lifelong-evolution capabilities.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Agent skill learning and evolution, one of 16 directions in LLM agents and tool use. Builds on 35 of its 51 references that are in the map. First posted on arXiv 18 Jul 2026.

Builds on

35 of 51 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026EvoSkill: Automated Skill Discovery for Multi-Agent Systems3 Mar
  2. 2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning9 Feb
  3. 2025Reinforcement Learning for Self-Improving Agent with Skill Library18 Dec

Other references · 32

  1. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  2. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,193
  3. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,081
  4. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  5. 2025A-MEM: Agentic Memory for LLM Agentscited by 1,000
  6. 2023ExpeL: LLM Agents Are Experiential Learnerscited by 814
  7. 2024Evaluating Very Long-Term Conversational Memory of LLM Agentscited by 800
  8. 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memorycited by 722
  9. 2023Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Modelscited by 551
  10. 2024Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Modelscited by 277
  11. 2024HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Modelscited by 273
  12. 2025ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memorycited by 253
20 more
  1. 2025BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agentcited by 183
  2. 2025GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Taskscited by 142
  3. 2025Memory OS of AI Agentcited by 140
  4. 2026MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agentscited by 129
  5. 2025Memento: Fine-tuning LLM Agents without Fine-tuning LLMscited by 119
  6. 2025G-Memory: Tracing Hierarchical Memory for Multi-Agent Systemscited by 112
  7. 2026AutoSkill: Experience-Driven Lifelong Learning via Skill Self-Evolutioncited by 105
  8. 2025MemGen: Weaving Generative Latent Memory for Self-Evolving Agentscited by 99
  9. 2025Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term Memorycited by 88
  10. 2023RecurrentGPT: Interactive Generation of (Arbitrarily) Long Textcited by 84
  11. 2023Think-in-Memory: Recalling and Post-thinking Enable LLMs with Long-Term Memorycited by 64
  12. 2026EverMemOS: A Self-Organizing Memory Operating System for Structured Long-Horizon Reasoningcited by 52
  13. 2026SkillNet: Create, Evaluate, and Connect AI Skillscited by 48
  14. 2024Compress to Impress: Unleashing the Potential of Compressive Memory in Real-World Long-Term Conversationscited by 36
  15. 2026CUA-Skill: Develop Skills for Computer Using Agentcited by 32
  16. 2026SkillFlow:Benchmarking Lifelong Skill Discovery and Evolution for Autonomous Agentscited by 19
  17. 2026RetroAgent: From Solving to Evolving via Retrospective Dual Intrinsic Feedbackcited by 16
  18. 2025SAMULE: Self-Learning Agents Enhanced by Multi-level Reflectioncited by 15
  19. 2026Agentic Proposing: Enhancing Large Language Model Reasoning via Compositional Skill Synthesiscited by 11
  20. 2026Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memorycited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026MemoryLake on MemoryArena: A Matched Study of Agent Memory Backendscited by 0
  2. 2026TTSE: A Two-Track Online Self-Evolution Frameworkcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026EvoAgentBench: Benchmarking Agent Self-Evolution via Ability Transfer
  2. 2026SkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles
  3. 2026Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills
  4. 2026Bayesian-Agent: Posterior-Guided Skill Evolution Across LLM Agent Harnesses
  5. 2026EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning
  6. 2026EVISKILL: Grounding Skill Evolution in Replayable Evidence
  7. 2026From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanisms
  8. 2026The Epistemics of Agent Memory: Measuring, and Governing, the Consolidation Decision in Long-Horizon LLM Agents