Harnessing AtomisticSkills for Agentic Atomistic Research

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Deng, Bowen, Li, Bohan, Cox, Matthew, Chun, Hoje, Nam, Juno, Lyssenko, Artur, Edamadaka, Sathya, Ruza, Jurgis, Du, Xiaochen, Segal, Nofit, Sanchez, Jesus Diaz, Xie, Mingrou, Perez, Ty, Yao, Yu, Steiner, Miguel, Majumdar, Sauradeep, Musgrave III, Charles B., Chandra, Anirban, Patra, Abhirup, Hohl, Detlef, Coley, Connor W., Li, Ju, Gómez-Bombarelli, Rafael
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911710295621632
author Deng, Bowen
Li, Bohan
Cox, Matthew
Chun, Hoje
Nam, Juno
Lyssenko, Artur
Edamadaka, Sathya
Ruza, Jurgis
Du, Xiaochen
Segal, Nofit
Sanchez, Jesus Diaz
Xie, Mingrou
Perez, Ty
Yao, Yu
Steiner, Miguel
Majumdar, Sauradeep
Musgrave III, Charles B.
Chandra, Anirban
Patra, Abhirup
Hohl, Detlef
Coley, Connor W.
Li, Ju
Gómez-Bombarelli, Rafael
author_facet Deng, Bowen
Li, Bohan
Cox, Matthew
Chun, Hoje
Nam, Juno
Lyssenko, Artur
Edamadaka, Sathya
Ruza, Jurgis
Du, Xiaochen
Segal, Nofit
Sanchez, Jesus Diaz
Xie, Mingrou
Perez, Ty
Yao, Yu
Steiner, Miguel
Majumdar, Sauradeep
Musgrave III, Charles B.
Chandra, Anirban
Patra, Abhirup
Hohl, Detlef
Coley, Connor W.
Li, Ju
Gómez-Bombarelli, Rafael
contents Computational materials science and chemistry span vast knowledge domains and fractured software ecosystems. Although large language models (LLMs) have demonstrated research capabilities, scaling monolithic agents to manage the rigor and complexity of atomistic research remains a challenge. Here, we introduce AtomisticSkills, an open-source harness framework that empowers general-purpose AI coding agents to conduct atomistic research across materials science, chemistry, and drug discovery. By hierarchically decomposing scientific workflows into agent skills and tools, AtomisticSkills provides agents with modular, extensible, and plug-and-play research capabilities. The framework integrates more than 100 human-curated multidisciplinary skills, including database access, thermodynamics and kinetics modeling, and diverse simulation engines employing machine learning interatomic potentials (MLIPs) and density functional theory (DFT). We validate its functional coverage against scientific literature and demonstrate robust orchestration capabilities across diverse scientific campaigns: generative design of Li-ion solid-state electrolytes, high-throughput screening of metal-organic frameworks for CO2 capture, autonomous MLIP benchmarking and fine-tuning, multi-stage structure-based virtual screening for drug design, multimodal X-ray diffraction pattern analysis, and screening of Fe-oxide catalysts for oxygen evolution reaction. AtomisticSkills provides a critical agent infrastructure towards building fully autonomous AI scientists.
format Preprint
id arxiv_https___arxiv_org_abs_2605_24002
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Harnessing AtomisticSkills for Agentic Atomistic Research
Deng, Bowen
Li, Bohan
Cox, Matthew
Chun, Hoje
Nam, Juno
Lyssenko, Artur
Edamadaka, Sathya
Ruza, Jurgis
Du, Xiaochen
Segal, Nofit
Sanchez, Jesus Diaz
Xie, Mingrou
Perez, Ty
Yao, Yu
Steiner, Miguel
Majumdar, Sauradeep
Musgrave III, Charles B.
Chandra, Anirban
Patra, Abhirup
Hohl, Detlef
Coley, Connor W.
Li, Ju
Gómez-Bombarelli, Rafael
Chemical Physics
Materials Science
Artificial Intelligence
Computational Physics
Computational materials science and chemistry span vast knowledge domains and fractured software ecosystems. Although large language models (LLMs) have demonstrated research capabilities, scaling monolithic agents to manage the rigor and complexity of atomistic research remains a challenge. Here, we introduce AtomisticSkills, an open-source harness framework that empowers general-purpose AI coding agents to conduct atomistic research across materials science, chemistry, and drug discovery. By hierarchically decomposing scientific workflows into agent skills and tools, AtomisticSkills provides agents with modular, extensible, and plug-and-play research capabilities. The framework integrates more than 100 human-curated multidisciplinary skills, including database access, thermodynamics and kinetics modeling, and diverse simulation engines employing machine learning interatomic potentials (MLIPs) and density functional theory (DFT). We validate its functional coverage against scientific literature and demonstrate robust orchestration capabilities across diverse scientific campaigns: generative design of Li-ion solid-state electrolytes, high-throughput screening of metal-organic frameworks for CO2 capture, autonomous MLIP benchmarking and fine-tuning, multi-stage structure-based virtual screening for drug design, multimodal X-ray diffraction pattern analysis, and screening of Fe-oxide catalysts for oxygen evolution reaction. AtomisticSkills provides a critical agent infrastructure towards building fully autonomous AI scientists.
title Harnessing AtomisticSkills for Agentic Atomistic Research
topic Chemical Physics
Materials Science
Artificial Intelligence
Computational Physics
url https://arxiv.org/abs/2605.24002