FlowEvo: Self-Evolving Agents through the Co-Evolution of Workflows and Executable Skills

Hugging Face Daily Papers Papers
llm-agents workflow-evolution skill-reuse inference-framework training-free benchmark-results open-source

Summary

FlowEvo is a training-free framework that enables large language model agents to co-evolve reusable skills and workflows at inference time, achieving state-of-the-art accuracy and efficiency across benchmarks like ALFWorld, HumanEval, and GSM8K.

Large language model agents can adapt to complex tasks by constructing workflows at inference time, but procedures discovered in one episode are usually discarded after execution. Existing skill libraries provide reusable executable routines, but are typically assembled offline and do not grow from the agent's own workflows. We introduce FlowEvo, a training-free framework in which workflows and skills co-evolve at inference time. FlowEvo compiles successful workflows into callable skills, stores them in a persistent bank, and uses retrieved skills either through direct execution or as context for constructing new workflows. It also tracks each skill's downstream utility and suppresses skills that cause negative transfer. Using a shared GPT-4o-mini backbone, FlowEvo achieves the highest accuracy among 8 baselines on the full standard splits of ALFWorld, HumanEval, MBPP, GSM8K, and MATH-500. On ALFWorld, it reaches 85.6%, 26.4 points above the strongest baseline, while using roughly one third as many tokens. Across 10 base models spanning 7B to 671B parameters, FlowEvo outperforms ExpeL in 49 of 50 model-dataset comparisons. Code is available at https://github.com/DEFENSE-SEU/FlowEvo.
Original Article
View Cached Full Text

Cached at: 08/22/26, 12:12 AM

Paper page - FlowEvo: Self-Evolving Agents through the Co-Evolution of Workflows and Executable Skills

Source: https://huggingface.co/papers/2607.21596 Published on Aug 20

·

Submitted byhttps://huggingface.co/LeoYML

Leo Yon Aug 21

Abstract

FlowEvo enables large language model agents to co-evolve reusable skills and workflows during inference, improving accuracy and efficiency across diverse benchmarks.

Large language model agents can adapt to complex tasks by constructing workflows at inference time, but procedures discovered in one episode are usually discarded after execution. Existingskill librariesprovide reusable executable routines, but are typically assembled offline and do not grow from the agent’s own workflows. We introduceFlowEvo, a training-free framework in which workflows and skills co-evolve at inference time.FlowEvocompiles successful workflows into callable skills, stores them in a persistent bank, and uses retrieved skills either through direct execution or as context for constructing new workflows. It also tracks each skill’s downstream utility and suppresses skills that causenegative transfer. Using a shared GPT-4o-mini backbone,FlowEvoachieves the highest accuracy among 8 baselines on the full standard splits of ALFWorld, HumanEval, MBPP, GSM8K, and MATH-500. On ALFWorld, it reaches 85.6%, 26.4 points above the strongest baseline, while using roughly one third as many tokens. Across 10 base models spanning 7B to 671B parameters,FlowEvooutperforms ExpeL in 49 of 50 model-dataset comparisons. Code is available at https://github.com/DEFENSE-SEU/FlowEvo.

View arXiv pageView PDFGitHub7Add to collection

Get this paper in your agent:

hf papers read 2607\.21596

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.21596 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.21596 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.21596 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

SkillFlow: Flow-Driven Recursive Skill Evolution for Agentic Orchestration

arXiv cs.AI

SkillFlow proposes a flow-driven recursive skill evolution framework for LLM-based agentic orchestration, using Tempered Trajectory Balance to prevent strategy collapse and provide transparent credit assignment. Experiments on 14 datasets show significant improvements over baselines in QA, math, code, and decision-making tasks.

SkillFlow:Benchmarking Lifelong Skill Discovery and Evolution for Autonomous Agents

Hugging Face Daily Papers

SkillFlow introduces a benchmark of 166 tasks across 20 families for evaluating autonomous agents' ability to discover, repair, and maintain skills over time through a lifelong learning protocol. Experiments reveal a substantial capability gap among leading models, with Claude Opus 4.6 improving significantly while others show limited or negative gains from skill evolution.

EvoMAS: Learning Execution-Time Workflows for Multi-Agent Systems

arXiv cs.AI

EvoMAS is a framework for learning execution-time workflows in multi-agent systems by formulating workflow construction as a sequential decision problem. It outperforms static multi-agent design methods on complex tasks by adapting agent coordination dynamically based on evolving task states.