mlsanityData to 7 Oct 2026

SkillMAS: Skill Co-Evolution with LLM-based Multi-Agent System

arXiv 2605.09341 ↗First posted 10 May 2026cs.MA cs.CLcited by 3

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

Abstract (arXiv)

Large language model (LLM) agent systems are increasingly expected to improve after deployment, but existing work often decouples two adaptation targets: skill evolution and multi-agent system (MAS) restructuring. This separation can create organization bottlenecks, context pressure, and mis-specialization. We present SkillMAS, a non-parametric framework for adaptive specialization in multi-agent systems that couples skill evolution with MAS restructuring. SkillMAS uses Utility Learning to assign credit from verified execution traces, bounded skill evolution to refine reusable procedures without unfiltered library growth, and evidence-gated MAS restructuring when retained failures and Executor Utility indicate a structural mismatch. Across embodied manipulation, command-line execution, and retail workflows, SkillMAS is competitive under the reported harnesses while clarifying how post-deployment specialization is attributed, updated, and applied.

Key facts. Cited by 3 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 36 of its 37 references that are in the map. First posted on arXiv 10 May 2026.

Builds on

36 of 37 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026How Well Do Agentic Skills Work in the Wild: Benchmarking LLM Skill Usage in Realistic Settings6 Apr
  2. 2026CoEvoSkills: Self-Evolving Agent Skills via Co-Evolutionary Verification2 Apr
  3. 2026Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills26 Mar
  4. 2026SkillRouter: Skill Routing for LLM Agents at Scale23 Mar
  5. 2026MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild17 Mar
  6. 2026EvoSkill: Automated Skill Discovery for Multi-Agent Systems3 Mar
  7. 2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning9 Feb

Other references · 29

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  3. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,081
  4. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
  5. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,761
  6. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,280
  7. 2020ALFWorld: Aligning Text and Embodied Environments for Interactive Learningcited by 1,143
  8. 2025ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memorycited by 253
  9. 2026MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agentscited by 129
  10. 2026MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memorycited by 113
  11. 2025MemEvolve: Meta-Evolution of Agent Memory Systemscited by 81
  12. 2026SWE-Skills-Bench: Do Agent Skills Actually Help in Real-World Software Engineering?cited by 72
17 more
  1. 2026Externalization in LLM Agents: A Unified Review of Memory, Skills, Protocols and Harness Engineeringcited by 68
  2. 2025LifelongAgentBench: Evaluating LLM Agents as Lifelong Learnerscited by 65
  3. 2025WebEvolver: Enhancing Web Agent Self-Improvement with Coevolving World Modelcited by 56
  4. 2025ReflAct: World-Grounded Decision Making in LLM Agents via Goal-State Reflectioncited by 30
  5. 2025CoMAS: Co-Evolving Multi-Agent Systems via Interaction Rewardscited by 27
  6. 2026Understanding Agent Scaling in LLM-Based Multi-Agent Systems via Diversitycited by 25
  7. 2025Self-Evolving Multi-Agent Systems via Textual Backpropagationcited by 21
  8. 2026SkillProbe: Security Auditing for Emerging Agent Skill Marketplaces via Multi-Agent Collaborationcited by 21
  9. 2025Self-Generated In-Context Examples Improve LLM Agents for Sequential Decision-Making Taskscited by 17
  10. 2024MorphAgent: Empowering Agents through Self-Evolving Profiles and Decentralized Collaborationcited by 15
  11. 2026AdaptOrch: Task-Adaptive Multi-Agent Orchestration in the Era of LLM Performance Convergencecited by 15
  12. 2025Position: The Real Barrier to LLM Agent Usability is Agentic ROIcited by 12
  13. 2025United Minds or Isolated Agents? Exploring Coordination of LLMs under Cognitive Load Theorycited by 10
  14. 2026SkillGraph: Self-Evolving Multi-Agent Collaboration with Multimodal Graph Topologycited by 8
  15. 2026AgentSpawn: Adaptive Multi-Agent Collaboration Through Dynamic Spawning for Long-Horizon Code Generationcited by 7
  16. 2026AutoAgent: Evolving Cognition and Elastic Memory Orchestration for Adaptive Agentscited by 5
  17. 2026Holos: A Web-Scale LLM-Based Multi-Agent System for the Agentic Webcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Skills on the Fly: Test-Time Adaptive Skill Synthesis for LLM Agentscited by 10
  2. 2026Demystifying Agent Skills for Smart Contract Auditing: Design, Effectiveness, Behavioral Impactcited by 0
  3. 2026EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignmentcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Skill-MAS: Evolving Meta-Skill for Automatic Multi-Agent Systems
  2. 2026SkillFocus: Evolving Agent Skills via Capability Decomposition
  3. 2026ContinualSkillBench: Can LLM Agents Truly Evolve Their Capabilities?
  4. 2026SkillClaw: Let Skills Evolve Collectively with Agentic Evolver
  5. 2026SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories
  6. 2026An Empirical Study of Downstream Adaptation for Agent Skills
  7. 2026SkillCommit: Evolving Agent Skills through Behaviorally Validated Scope Expansion
  8. 2026Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward