MetaSkill-Evolve: Recursive Self-Improvement of LLM Agents via Two-Timescale Meta-Skill Evolution

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Zefeng Wang, Minxi Yan, Jinhe Bi, Sikuan Yan, Volker Tresp, Yunpu Ma
• Venue: arXiv: cs.AI
• Published: 2026-07-06
• Collected: 2026-07-07
• URL: http://arxiv.org/abs/2607.05297v1
• PDF: https://arxiv.org/pdf/2607.05297v1
• DOI: UNKNOWN

Plain-Language Brief

MetaSkill-Evolve is a framework that allows for recursive self-improvement of Large Language Model (LLM) agents. This is achieved by introducing a two-timescale system where each agent has both a task skill and a branch-local meta-skill. The meta-skill includes five components that control the improvement pipeline. On three agentic benchmarks, MetaSkill-Evolve outperformed baselines, improving test accuracy over the raw backbone.

What the Document Says

• MetaSkill-Evolve is a framework for recursive self-improvement of LLM agents.
• It introduces a two-timescale system where each agent has both a task skill and a branch-local meta-skill.
• The meta-skill includes five components: Analyzer, Retriever, Allocator, Proposer, and Evolver.
• Task skills evolve on a fast loop, while the meta-skill evolves on a slower one.
• MetaSkill-Evolve outperformed no-skill, static-skill, and single-level evolution baselines on three agentic benchmarks: OfficeQA, SealQA, ALFWorld.
• It improved held-out test accuracy over the raw backbone by +23.54, +16.09, and +1.92 points respectively.

Key Concepts Mentioned

• MetaSkill-Evolve
• Large Language Model (LLM) agents
• two-timescale system
• task skill
• branch-local meta-skill
• Analyzer
• Retriever
• Allocator
• Proposer
• Evolver

Claims Made by the Document

• MetaSkill-Evolve allows for recursive self-improvement of LLM agents.
• It outperformed no-skill, static-skill, and single-level evolution baselines on three agentic benchmarks.

Evidence Used

• Improvement is made possible through the two-timescale system and the five components of the meta-skill.
• The performance improvements over the baselines are stated in the text.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Recent LLM agents tackle increasingly long-horizon, open-ended tasks, and external skills, reusable procedural knowledge supplied to the agent, further extend this capability. However, a fixed, hand-authored skill is rarely optimal, and cannot adapt to the diversity of tasks an agent encounters. Self-improving agents address this by rewriting their own skill files from execution traces, yielding meaningful gains on challenging benchmarks. Yet such self-evolution remains non-recursive: it improves only the task skill (what the agent does) while the improvement procedure (how it improves) is authored once and held fixed. We introduce MetaSkill-Evolve, a two-timescale framework that makes agentic skill improvement recursive: every branch carries both a task skill $s$ and a branch-local meta-skill $m=(ψ,σ,α,π,\varepsilon)$ whose five components parameterise the Analyzer, Retriever, Allocator, Proposer, and Evolver agents of the improvement pipeline. Task skills evolve on a fast loop while the meta-skill evolves on a slower one under the same pipeline applied to itself, with no additional model or objective. With all five pipeline agents sharing a single frozen backbone, MetaSkill-Evolve outperforms no-skill, static-skill, and single-level evolution baselines on three agentic benchmarks (OfficeQA, SealQA, ALFWorld), improving held-out test accuracy over the raw backbone by +23.54, +16.09, and +1.92 points respectively.