Model Context Protocol (MCP)
Browse MCP servers you can connect. Probe tools, then enable only what an agent or ActionFlow should see.
Frequently asked questions
Model Context Protocol (MCP) is an open protocol for exposing tools to AI clients. In ActionFlows, MCP servers are catalog entries you connect so an agent or ActionFlow can call those tools.
It is a standard way to attach tools without a one-off integration for every host. You connect a server, probe the tools, and enable only what that agent or ActionFlow should see.
Open the MCP catalog, connect a server with the required credentials or OAuth, then attach it on an agent in Studio, or use it from the tools that support it. Unconfigured servers are skipped and do not block the rest of the setup.
You grant only the tools you enable. Organization credentials stay on the organization. Agent output can use personal data (PII) redaction and moderation. MCP does not by itself encrypt every system you connect. Treat those servers as production access.
Start building AI workflows
Create a free account, open a template or a blank canvas, and run your first ActionFlow.
Get product updates
New nodes, agents, and product notes. We send mail only when we have something worth opening.