LabVIEW MCP servers and AI agent skills

How AI connects to LabVIEW

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

• VI Scripting and VI Server (Other): VI Scripting lets a LabVIEW program create VIs, add and wire objects and move or delete them through VI Server properties and methods; NI's Common VI Scripting Tasks page maps each job to the methods involved. The CalmyJane server uses it from a DQMH module running in LabVIEW, and Zühlke's server falls back to the ActiveX VI Server on versions before 2026 Q3. Documentation
• Python Node (Python): The Python Node, added in LabVIEW 2018, calls Python scripts from a block diagram and, since 2023 Q1, supports Anaconda and venv environments. It runs the other way from an MCP server (LabVIEW calling Python), but it is how a VI can reuse Python analysis code that an agent writes. Documentation

What NI offers

NI's Nigel AI is built into LabVIEW. In LabVIEW 2026 Q3 it can generate measurement VIs from a prompt for connected NI hardware, answer questions about projects, targets and installed NI software, let you pick the model, and connect to MCP servers and tools as a client. Nigel runs on Microsoft Azure OpenAI. NI publishes no MCP server that exposes LabVIEW to outside clients.

Checked 4 Nov 2026. Sources: NI Nigel AI capabilities in LabVIEW (NI documentation), Generating VIs with NI Nigel AI (NI documentation)

What you can do with LabVIEW and AI

• Explain what an existing VI does from its block diagram returned as text, or summarise a whole .lvproj or .lvlib.
• Generate a new VI from a description, validate it, save it as a real .vi file and run it to check the outputs.
• Add error handling or other changes to an existing VI in place, working on a copy under version control.
• Search installed palettes and NI's shipping examples for an existing VI before writing a new one.
• Document a LabVIEW library, class or project as a Word file with one section per public VI.

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, Cursor and GitHub Copilot. Install steps, copied from each project's own documentation, are on each listing's page. Most LabVIEW integrations run on your own computer next to LabVIEW, so you need LabVIEW installed and licensed on the same machine unless the listing says otherwise.