---
title: "From Atomic Actions to Standard Operating Procedures: Iterative Tool Optimization for Self-Evolving LLM Agents"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.07321v1"
permalink: "https://feed.astralor.com/items/itm_1ff7cb29e11e18902b148a4d"
published: "2026-07-08T12:09:49.000Z"
score: 78
channel: "tech"
tags: ["EvoSOP","LLM Agent","Tool Optimization"]
---

# From Atomic Actions to Standard Operating Procedures: Iterative Tool Optimization for Self-Evolving LLM Agents

## AI commentary

> This practical research on self-evolving agents addresses the core pain point of static toolsets. It is especially valuable for developers working on autonomous agents and tool learning, offering a complete iterative framework.

## At a glance

This study introduces EvoSOP, a framework that enables LLM agents to synthesize atomic actions (e.g., basic file I/O, single-turn search) into reusable Standard Operating Procedures (SOPs). Through a systematic lifecycle of construction, merging, evaluation, and pruning, agents iteratively optimize their toolset. Extensive experiments show significant improvements in task success rates and reduced interaction rounds, offering a scalable pathway for self-evolving agents.

## Key points

- Proposes EvoSOP for extracting SOPs as higher-order tools from execution trajectories
- Includes a full lifecycle of construction, merging, evaluation, and pruning
- Experiments show significant improvement in task success rate and fewer interaction rounds

## Full text

Tool utilization enables Large Language Model (LLM) agents to interact with the real world and resolve complex tasks. However, existing agent frameworks predominantly rely on static toolsets composed of granular atomic actions (e.g., basic file I/O or single-turn search), which forces agents to reinvent low-level logic for every recurring workflow, leading to increased reasoning overhead and failure rates. In this study, we propose that agents can achieve self-evolution by synthesizing these atomic actions into reusable Standard Operating Procedures (SOPs), which function as callable higher-order tools that encapsulate multi-step logic. We further introduce EvoSOP, a framework that empowers agents to extract SOPs from execution trajectories and iteratively optimize the toolset through a systematic lifecycle of construction, merging, evaluation, and pruning. Extensive experiments demonstrate that EvoSOP significantly boosts task success rates while substantially reducing the number of interaction rounds compared to baselines. Our analysis also reveals that iterative tool optimization fosters reliable and efficient tool-use patterns, providing a scalable pathway for the development of self-evolving agents.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_1ff7cb29e11e18902b148a4d