Matlantis announced a new AI agent integration for its atomistic simulation platform that allows researchers to create, edit, and run materials simulations using natural-language instructions. The release includes a public Skills library on GitHub and a forthcoming installer that will allow Anthropic’s Claude Code to run directly within the Matlantis terminal environment.
The platform is intended to reduce the programming and scripting expertise traditionally required for atomistic simulation workflows used in materials research, including battery development for electric vehicle (EV) and energy storage applications.
Atomistic simulation has historically required expertise in computational chemistry, programming, and software environments, limiting its use primarily to specialists. Matlantis previously introduced its AI-based Preferred Potential (PFP) model as a cloud service to reduce infrastructure complexity. The latest release focuses on simplifying the scripting and workflow layer between a research objective and a working simulation.
By embedding AI agents directly into the simulation environment, researchers can generate, edit, and run simulations through conversational prompts rather than manual scripting. The public Skills library packages Matlantis-specific APIs, functions, and workflows into a format accessible to AI agents, enabling more context-aware simulation generation.
Initial workflows supported by the library include structure relaxation, molecular dynamics, reaction pathway exploration, crystal structure prediction, visualization, and retrieval of structures from external databases.
A forthcoming update will allow users to launch Claude Code directly from the Matlantis terminal environment. Researchers will be able to describe a simulation objective in natural language, generate or modify scripts, run calculations, and interpret results within the same workflow environment.
The integration is intended to help experimental researchers without programming backgrounds access simulation tools more easily, while allowing computational specialists to reduce repetitive scripting tasks and accelerate candidate screening and workflow adaptation.
Matlantis said the platform is designed to support broader adoption of computational materials research across industrial R&D applications, including battery materials development and other advanced materials workflows.