Developers aren’t only working on platform consoles anymore. They’re building in Claude Desktop, Windsurf, Cursor, VS Code, and other AI-assisted tools. Platform teams are troubleshooting in Slack. AI agents are starting to call APIs, use MCP servers, and interact with enterprise systems directly.
Headless MuleSoft is built for that reality. We are now giving developers, platform teams, and agents a way to access MuleSoft from the tools where they already work.
The Platform MCP Server brings MuleSoft inside your AI workflow
Introducing the MuleSoft Platform MCP Server, a remote MCP server that allows any MCP-compatible client, like Claude Desktop, Windsurf, Cursor, VS Code or your own agent, to connect directly to MuleSoft. Once connected, developers and agents can work with their APIs, agents, MCP servers, LLMs, governance data, and monitoring information without switching to the Anypoint UI. That means a developer can ask their AI tool questions like: “Which APIs do we already have for customer onboarding?” or “Show me the governance status for this API,” without leaving the AI tool they are already using.
See it in action
Headless MuleSoft goes beyond providing convenience for the user. It reduces the gap between development and governance. Teams can discover, use, and manage MuleSoft assets from AI-native tools through MuleSoft’s permission model and platform controls. The experience changes, the governance doesn’t. Watch how this all comes together in Claude Desktop in our demo:
Looking ahead
Headless MuleSoft is the new operating model for organizations building headless enterprises with MuleSoft. MuleSoft platform capabilities are now available from wherever your team already works, with the same governance model regardless of how they’re accessed. Learn more about MuleSoft Platform MCP Server in the full documentation.
