Introduction
PowerAgent is an open-source community dedicated to accelerating the development of LLM-powered Tools and Agentic AI in the power systems domain. We aim to bridge the gap between cutting-edge AI and the real-world needs of system operators, electric utilities, and researchers, by creating a shared ecosystem of protocols, models, and workflows tailored for power applications.
We believe the future of agentic AI in power domain will be driven by the combined development and coordination of three key components — Foundation Models (FM), Model Context Protocol (MCP), and Workflows (WF).
PowerFM
Fine-tuned LLMs for power system tasks, full domain-specific Foundation Models, and benchmarking for model performance
PowerMCP
Protocols enabling LLMs to control and interact with power domain software and tools
PowerWF
Modular agentic workflows for orchestrating models, tools, and decisions
Motivation
The rapid rise of LLMs and agentic frameworks has opened new frontiers in intelligent software. Inspired by the success of open-source initiatives in LLMs (like DeepSeek and Llama) and agentic control protocols (especially Anthropic's MCP community), we believe it's time to bring the same spirit of innovation and openness to the power systems domain.
Vision
Toward General Intelligence for Power Grids through Open-Source Community
We are building a comprehensive ecosystem based on three interconnected layers: workflow automation, tool integration, and domain-specialized foundation models. Through this layered architecture, we aim to deliver scalable and intelligent solutions, accelerate the deployment of agentic AI systems, and foster an open, collaborative community within the power sector.
Demos
Contingency Evaluation
An LLM automatically operates power system software, such as PowerWorld and pandapower, to perform contingency analysis and generate professional reports.
Load Growth Evaluation
An LLM automatically operates power system software, such as PowerWorld, to evaluate different load growth scenarios and generate professional reports with recommendations.
Electricity Market Expert
An automated workflow that handles electricity market inquiries, generates professional responses using an LLM, logs all actions for traceability, and supports human-in-the-loop review.
Datacenter Siting Expert
Datacenter Siting Assistant: Solvtra is a tool leverages RAG by incorporating datacenter-specific data, including local regulations, environmental reports, and infrastructure details. As a result, it can provide detailed information for potential datacenter locations, such as land and electricity prices, and show a map illustrating existing datacenter sites and relevant infrastructure. Try it on Hugging Face Spaces.
Community
Like other open-source communities, PowerAgent is built by and for the community. GitHub was chosen as the home of this initiative to enable transparent, global collaboration. Maintained by the Power and AI Initiative (PAI) at Harvard SEAS, all contributions are reviewed to ensure they are relevant and open-source, utility, and integrity. We invite researchers, engineers, and developers across disciplines to take part in shaping this ecosystem.
Simple and Unrestricted Access
To keep PowerAgent accessible with minimal effort, all current benchmarks are maintained and tested using Cursor, one of the most popular AI coding IDEs. In the interest of open collaboration and replicable experiments, all PowerAgent code are distributed under the MIT license. While not required, we appreciate credit or acknowledgment of the original creators when reusing or building upon these resources.
Core Team
Le Xie, Qian Zhang, Shaohuai Liu, Lin Dong, Alban Puech
For academic use, please cite the following paper:
Qian Zhang, Le Xie. "PowerAgent: A Roadmap Towards Agentic Intelligence in Power Systems", working paper, 2025.
How to Contribute
PowerAgent is a collaborative community-driven initiative. PowerAgent users are encouraged to fork benchmark repositories and modify them for their own purposes and then contribute their enhancements back to the broader community via pull requests .
To join the PowerAgent community core team or ask questions about contributions, please reach out to the administrator: Qian Zhang
Stay Updated
Subscribe to our newsletter for updates and announcements