EvolveTool-Bench: Evaluating the Quality of LLM-Generated Tool Libraries as Software Artifacts

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kaliyev, Alibek T., Maryanskyy, Artem
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914435991339008
author Kaliyev, Alibek T.
Maryanskyy, Artem
author_facet Kaliyev, Alibek T.
Maryanskyy, Artem
contents Modern LLM agents increasingly create their own tools at runtime -- from Python functions to API clients -- yet existing benchmarks evaluate them almost exclusively by downstream task completion. This is analogous to judging a software engineer only by whether their code runs, ignoring redundancy, regression, and safety. We introduce EvolveTool-Bench, a diagnostic benchmark for LLM-generated tool libraries in software engineering workflows. Across three domains requiring actual tool execution (proprietary data formats, API orchestration, and numerical computation), we define library-level software quality metrics -- reuse, redundancy, composition success, regression stability, and safety -- alongside a per-tool Tool Quality Score measuring correctness, robustness, generality, and code quality. In the first head-to-head comparison of code-level and strategy-level tool evolution (ARISE vs. EvoSkill vs. one-shot baselines, 99 tasks, two models), we show that systems with similar task completion (63-68%) differ by up to 18% in library health, revealing software quality risks invisible to task-only evaluation. Our results highlight that evaluation and governance of LLM-generated tools require treating the evolving tool library as a first-class software artifact, not a black box.
format Preprint
id arxiv_https___arxiv_org_abs_2604_00392
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle EvolveTool-Bench: Evaluating the Quality of LLM-Generated Tool Libraries as Software Artifacts
Kaliyev, Alibek T.
Maryanskyy, Artem
Software Engineering
Artificial Intelligence
Modern LLM agents increasingly create their own tools at runtime -- from Python functions to API clients -- yet existing benchmarks evaluate them almost exclusively by downstream task completion. This is analogous to judging a software engineer only by whether their code runs, ignoring redundancy, regression, and safety. We introduce EvolveTool-Bench, a diagnostic benchmark for LLM-generated tool libraries in software engineering workflows. Across three domains requiring actual tool execution (proprietary data formats, API orchestration, and numerical computation), we define library-level software quality metrics -- reuse, redundancy, composition success, regression stability, and safety -- alongside a per-tool Tool Quality Score measuring correctness, robustness, generality, and code quality. In the first head-to-head comparison of code-level and strategy-level tool evolution (ARISE vs. EvoSkill vs. one-shot baselines, 99 tasks, two models), we show that systems with similar task completion (63-68%) differ by up to 18% in library health, revealing software quality risks invisible to task-only evaluation. Our results highlight that evaluation and governance of LLM-generated tools require treating the evolving tool library as a first-class software artifact, not a black box.
title EvolveTool-Bench: Evaluating the Quality of LLM-Generated Tool Libraries as Software Artifacts
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2604.00392