Skip to content
Tresca

OpenDSS MCP servers and AI agent skills

Three community MCP servers run EPRI's OpenDSS distribution simulator in-process: PowerMCP's OpenDSS server (built on py-dss-toolkit and py-dss-interface), a standalone server with IEEE test feeders, hosting capacity and DER siting, and a newer one with fault, Thevenin and QSTS tools. A PowerSkills skill covers OpenDSS feeder studies. None needs a licence, and the py-dss-interface engine runs on Windows and Linux. EPRI publishes no AI tooling for OpenDSS.

  • No official EPRI (open source) server yet.
  • 3 MCP servers and 1 agent skill listed.
EPRI (open source)3 servers · 1 skillChecked 4 Nov 2026

Integrations for OpenDSS

Community and commercial servers

MCP serverCommunity

PowerMCPby Power-Agent

One Python package with separate MCP servers for pandapower, PyPSA, OpenDSS, PowerWorld, PSS/E, PowerFactory, PSCAD and more.

LocalWindows, macOS, LinuxNeeds a OpenDSS licenceCan modify models235· updated 8 days ago
MCP serverCommunity

OpenDSS MCP Server (jpsalamanca-co)by jpsalamanca-co

Twenty typed OpenDSS tools for power flow, violations, faults, Thevenin impedance, hosting capacity and QSTS runs.

LocalWindows, LinuxCan modify models1· updated 6 weeks ago
MCP serverCommunity

OpenDSS MCP Serverby ahmedelshazly27

Runs OpenDSS distribution studies from chat: IEEE test feeders, power flow, hosting capacity, DER siting and plots.

LocalWindows, macOS, LinuxCan modify models4· updated 1 year ago

Agent skills

SkillCommunity

PowerSkillsby Power-Agent

Skills that tell an agent which PowerMCP tool to use first, plus ten mitigation playbooks for power-system study findings.

80· updated 2 weeks ago

Listings are sorted by origin, then by recent activity. GitHub data is refreshed nightly. Nothing here is ranked.

How AI connects to OpenDSS

An MCP server or skill can only do what OpenDSS lets programs do. OpenDSS exposes 2 automation interfaces:

  • COM server and direct DLL (COM automation): The OpenDSS DLL registers as an in-process COM server on Windows, and on platforms without COM a direct DLL duplicates the same interface with plain function calls. EPRI says users commonly drive it from MATLAB, Python, C# and R. Every command of the text scripting language is also available through these interfaces, so an agent can compile a master .dss file and then query elements. Documentation
  • Python packages (py-dss-interface, py-dss-toolkit) (Python): py-dss-interface bundles an OpenDSS engine with Python bindings, and py-dss-toolkit adds model records and result tables on top. The PowerMCP and jpsalamanca-co servers use them, which is why neither needs a separate OpenDSS install. Paulo Radatz, who maintains py-dss-interface, is also on the PowerMCP core team, and PowerMCP's OpenDSS tools return the toolkit's element records as JSON. Documentation

What EPRI (open source) offers

EPRI's OpenDSS documentation describes the COM server, direct DLL and text scripting interface and says nothing about AI assistants, LLMs or MCP. OpenDSS is EPRI's open-source distribution simulator, so the integrations that exist are community projects layered on its Python interfaces, with no EPRI-run server.

Checked 4 Nov 2026. Sources: OpenDSS documentation (EPRI)

What you can do with OpenDSS and AI

  • Compile a master .dss feeder, solve a snapshot power flow and list buses outside the ANSI C84.1 voltage range.
  • Add PV or storage step by step until a voltage, thermal or convergence limit binds, to estimate hosting capacity on a feeder.
  • Run three-phase and single-line-to-ground faults across buses and return positive- and zero-sequence Thevenin impedances.
  • Run daily or yearly quasi-static time series with the circuit's load shapes and summarise losses and regulator behaviour.
  • Edit, add or disable elements such as loads, PV systems and regulators, then re-solve and compare the voltage profile.

Taken from the listings' own documentation. Check each listing for exactly what it supports.

Installing

The listings here document setup for Claude Code, Claude Desktop, Codex CLI and Cursor. Install steps, copied from each project's own documentation, are on each listing's page. Most OpenDSS integrations run on your own computer next to OpenDSS, so you need OpenDSS installed and licensed on the same machine unless the listing says otherwise.

Related tools

pandapowerOpen source (University of Kassel and Fraunhofer IEE)1 server · 1 skillCYMEEatonNone listed yetPowerFactoryDIgSILENT2 servers · 1 skillPyPSAOpen source (TU Berlin)2 servers · 1 skillPowerWorld SimulatorPowerWorld1 server · 1 skill

Questions

Is there an official OpenDSS MCP server?

No. EPRI's OpenDSS documentation doesn't mention AI or MCP. All three OpenDSS servers here are community projects, one of them part of PowerMCP.

Do I need to install OpenDSS separately?

Usually not. The servers use py-dss-interface, which ships the OpenDSS engine with its Python package, so a pip install is enough on Windows or Linux.

Which OpenDSS server should I start with for hosting capacity?

Both standalone servers have hosting-capacity tools. The jpsalamanca-co server returns typed results with the inputs that produced them; the ahmedelshazly27 server adds DER placement search, batch scenarios and IEEE test feeders.

AI-generated models, drawings and results must be checked by a qualified engineer before use.

Sources (checked 4 Nov 2026): OpenDSS documentation (EPRI) · PowerMCP OpenDSS server README · jpsalamanca-co/opendss-mcp README · ahmedelshazly27/opendss-mcp-server README · OpenDSS product page

OpenDSS is developed by its own project (EPRI (open source)). Tresca is independent and not affiliated with or endorsed by the OpenDSS project.