• AWS Open Source Blog Open Protocols for Agent Interoperability Part 2: Authentication on MCP In Part 1 of our blog series on Open Protocols for Agent Interoperability we covered how the Model Context Protocol (MCP) can be used to facilitate inter-agent communication and the MCP specification enhancements AWS is working on to enable that. • In Part 2 of this blog series we dive deep into authentication in the latest version of the MCP specification and discuss some of the contributions from AWS in this release. • The Model Context Protocol (MCP), created by Anthropic, has seen remarkable adoption since its November 2024 launch, drawing interest from developers and organizations worldwide. • Initially, MCP kept things straightforward - users simply downloaded and ran local MCP servers right on their workstations. • In March, MCP formalized its approach to remote server communication using the Streamable HTTP paradigm. • These remote servers eliminate the need for local software installation and updates, reducing security risks and deployment complexity while ensuring users always access the latest version of the service.
Article Summaries:
- AWS’s Open Source Blog released Part 2 of its “Open Protocols for Agent Interoperability” series, detailing authentication updates to the Model Context Protocol (MCP). MCP, originally launched by Anthropic in November 2024, has shifted from local servers to remote, stream‑based HTTP endpoints, raising the need for secure access. AWS has collaborated with MCP contributors to fill authentication gaps, adding a comprehensive OAuth‑based approach in the June 18 2025 specification release. The update includes a Java SDK pull request that implements the new authentication logic, enabling authenticated remote hosting of MCP servers on AWS and other cloud platforms.
Sources: