PTC sets out when to adopt, extend or build MCP connections for Windchill
A PTC blog post by Meg Folcarelli gives customers a simple framework for connecting AI tools to PTC products through the Model Context Protocol (MCP), an open standard that lets an AI application discover and use the capabilities of enterprise software. PTC's advice is to adopt its own supported MCP capabilities where they exist, extend them for specific workflows and integrations, and build only where that creates real business value. The argument is that access to data is the easy part. An agent also needs product context, such as revisions, configurations, structures, lifecycle states, requirements and change objects, or it may find the right data and still pick the wrong revision. PTC stresses that MCP is not itself the governance layer: Windchill remains responsible for permissions, business rules and audit. The post says PTC-supported MCP capabilities are intended for Windchill, and names an Onshape FeatureScript MCP server as a second example. It adds that availability will depend on the product and release, so customers should check before planning around it.
Sources
PTC, PLM blog: Adopt, Extend, or Build: Where MCP Creates the Most Value
Labelled “Vendor announcement” because the facts come from the vendor’s own announcement or results. This is not paid content. Claims are the vendor’s own unless stated.