ABACUS
ABACUS, an open-source DFT-based simulation platform, is supported by an MCP server that provides a structured communication interface for submitting, managing, and executing tasks.
What is ABACUS?
ABACUS is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to abacus, an open-source dft-based simulation platform, is supported by an mcp server that provides a structured communication interface for submitting, managing, and executing tasks.
ABACUS, an open-source DFT-based simulation platform, is supported by an MCP server that provides a structured communication interface for submitting, managing, and executing tasks.
This server falls under the Data Science & ML category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- ABACUS, an open-source DFT-based simulation platform, is sup
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx abacusConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use ABACUS
The ABACUS MCP Server connects AI assistants to the ABACUS open-source Density Functional Theory (DFT) simulation platform, enabling automated submission and management of computational chemistry and materials science calculations. Through structured MCP tools, users can create crystal structures, run SCF, geometry optimization, and molecular dynamics tasks, then retrieve band structures, DOS, and charge density results — all from within an AI conversation.
Prerequisites
- Python 3.8 or higher installed
- ABACUS software package installed and accessible via the abacus command
- Pseudopotential files (UPF format) in a local pseudos/ directory
- Sufficient computational resources (CPU cores, memory) for DFT calculations
- Git to clone the repository
Clone the repository and install Python dependencies
Clone the abacus-mcp-server repository and install the required Python packages including FastMCP, ASE, and NumPy.
git clone https://github.com/PhelanShao/abacus-mcp-server.git
cd abacus-mcp-server
pip install -r requirements.txt
pip install fastmcp ase numpy psutilVerify ABACUS is accessible
Confirm the ABACUS executable is on your PATH and returns a version string. This is required before the MCP server can submit calculations.
abacus --versionConfigure environment variables
Set the required environment variables pointing to your ABACUS executable, pseudopotential files directory, and working directory for calculation outputs.
export ABACUS_COMMAND=abacus
export PSEUDO_DIR=/path/to/pseudos
export ABACUS_WORK_DIR=/path/to/calculations
export LOG_LEVEL=INFO
# Or create a config.json:
# {
# "abacus_command": "abacus",
# "pseudo_base_path": "./pseudos",
# "work_directory": "./calculations"
# }Start the MCP server
Launch the server by running the main server script. It will start in stdio mode for direct MCP client integration.
python src/server.pyAdd to Claude Desktop configuration
Register the ABACUS MCP server in your Claude Desktop config file so it launches automatically when Claude starts.
{
"mcpServers": {
"abacus": {
"command": "python",
"args": ["/path/to/abacus-mcp-server/src/server.py"],
"env": {
"ABACUS_COMMAND": "abacus",
"PSEUDO_DIR": "/path/to/pseudos",
"ABACUS_WORK_DIR": "/path/to/calculations",
"LOG_LEVEL": "INFO"
}
}
}
}ABACUS Examples
Client configuration
Claude Desktop configuration for the ABACUS MCP Server with required environment variables for paths and the ABACUS executable.
{
"mcpServers": {
"abacus": {
"command": "python",
"args": ["/home/user/abacus-mcp-server/src/server.py"],
"env": {
"ABACUS_COMMAND": "abacus",
"PSEUDO_DIR": "/home/user/pseudos",
"ABACUS_WORK_DIR": "/home/user/calculations",
"LOG_LEVEL": "INFO"
}
}
}
}Prompts to try
Example prompts for running DFT calculations with the ABACUS MCP server.
- "Create a diamond silicon structure with lattice constant 5.43 Angstrom and validate the input parameters"
- "Run an SCF calculation for silicon using a 4x4x4 Monkhorst-Pack k-point grid and 100 Ry cutoff energy"
- "Calculate the band structure of my optimized silicon structure along the G-X-M-G path"
- "Check the status of my last calculation and retrieve the total energy result"
- "Suggest appropriate DFT parameters for a geometry optimization of bulk iron"Troubleshooting ABACUS
Server starts but abacus command is not found when submitting calculations
Verify ABACUS is installed and on the system PATH by running abacus --version in a terminal. Set the full absolute path in the ABACUS_COMMAND environment variable (e.g., ABACUS_COMMAND=/usr/local/bin/abacus) to avoid PATH issues when launched from Claude Desktop.
SCF calculation fails with pseudopotential not found error
Ensure UPF pseudopotential files for all elements in your structure are present in the PSEUDO_DIR directory. The pseudo_potential_map parameter in tool calls must match the exact filenames in that directory (e.g., Si.UPF for silicon).
pip install fails for fastmcp or ASE packages
Create a virtual environment to avoid conflicts: python -m venv venv && source venv/bin/activate, then run pip install -r requirements.txt. Ensure Python 3.8+ is active with python --version before installing.
Frequently Asked Questions about ABACUS
What is ABACUS?
ABACUS is a Model Context Protocol (MCP) server that abacus, an open-source dft-based simulation platform, is supported by an mcp server that provides a structured communication interface for submitting, managing, and executing tasks. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install ABACUS?
Follow the installation instructions on the ABACUS GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with ABACUS?
ABACUS works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is ABACUS free to use?
Yes, ABACUS is open source and available under the GPL-3.0 license. You can use it freely in both personal and commercial projects.
ABACUS Alternatives — Similar Data Science & ML Servers
Looking for alternatives to ABACUS? Here are other popular data science & ml servers you can use with Claude, Cursor, and VS Code.
Ultrarag
★ 5.6kA Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines
RocketRide
★ 3.1k📇 🏠 - MCP server that exposes RocketRide AI pipelines as t
Aix Db
★ 2.1kAix-DB 基于 LangChain/LangGraph 框架，结合 MCP Skills 多智能体协作架构，实现自然语言到数据洞察的端到端转换。
NeMo Data Designer
★ 1.9k🎨 NeMo Data Designer: Generate high-quality synthetic data from scratch or from seed data.
PaperBanana
★ 1.7kOpen source implementation and extension of Google Research’s PaperBanana for automated academic figures, diagrams, and research visuals, expanded to new domains like slide generation.
MiniMax
★ 1.5kBridges MiniMax AI capabilities to the Model Context Protocol, enabling AI agents to perform image understanding, text-to-image generation, and speech synthesis. It provides a standardized interface for accessing MiniMax's core tools via JSON-RPC.
Browse More Data Science & ML MCP Servers
Explore all data science & ml servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up ABACUS in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use ABACUS?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.