Agentic messaging is next on the to-do list for the team behind agentic AI protocol MCP, one of its co-creators has said during a keynote at the Linux Foundation’s first MCPCon in Europe.
Anthropic Technical Staff Member and MCP Co-Creator David Soria Parra said the protocol’s maintainers were expanding the “MCP tasks” extension to enable long-running operations through the protocol as personal and multiagent use cases increased.
He told the audience in Amsterdam that MCP was increasingly seeing that “people don’t just want to call tools, but they want to call effectively full-on agents from an MCP client… so we really want to push for this type of agentic messaging in MCP and make it the default protocol for connecting clients with agents.”
Future releases will also focus on adding richer semantics and helping users deal with authorisation and identity issues for agents running in production, he said.
