Get a head start on AI agent workflows with open-source skills

We see many of you building custom AI agents and skills to automate and streamline your architectural work by connecting to your workspace via the MCP server. To get those of you who haven't done so yet started and inspire everybody, we've launched an open-source repository to build on.

It comes preloaded with a curated set of skills that teach any AI agent how to perform four core workspace operations: search the inventory in plain language, run calculations across your landscape, build and debug automations, and generate architecture diagrams from a short description. Each skill package contains the knowledge agents need to complete the task with the respective MCP tools, ensuring reliable results rather than approximations.

Because the skills are open source, you can see exactly what an agent will do before it runs, adapt any skill to your own governance rules, and use the same skills across the AI tools your teams already work in. There's no proprietary SDK and no integration code to maintain. This builds on the open-source plugins we announced earlier this year, and works with the MCP server and AI Agent Hub you already use to connect and govern your agents.

For an overview of the available skills, check out the repository's README.