MCP and Agent Skills standards join Linux Foundation's Agentic AI Foundation

The Model Context Protocol (MCP) and the Agent Skills specification have officially joined the Linux Foundation's Agentic AI Foundation (AAIF). This transition follows Anthropic's donation of the Agent Skills specification to the foundation, moving these critical AI standards toward vendor-neutral governance. The move aims to prevent ecosystem fragmentation by providing a shared framework for how AI agents interact with external data and tools.

Engineers from major industry players, including Google, Anthropic, and OpenAI, are now co-maintaining these standards. This collaborative approach is designed to solve the 'N-by-M' integration problem, where developers previously had to write custom connectors for every combination of AI agent framework and data source. By standardizing these interfaces, the industry seeks to build more scalable, production-ready AI agents.

Technical evolution of MCP and Agent Skills

• Model Context Protocol (MCP): Provides a standardized method for connecting AI agents to live external tools and data sources, reducing the need for custom integrations between clients and hosts.
• Agent Skills Specification: Enables agents to load procedural domain knowledge, including instructions and scripts, on demand to manage complex tasks without exhausting the model's context window.
• Stateless Architecture: The MCP specification was updated on July 28, 2026, to transition from a stateful model to a stateless, horizontally scalable architecture to support production-scale deployments.
• MCP Toolbox for Databases: An open-source framework that implements MCP to standardize how agents look up and invoke tools across various database types, addressing scaling challenges in cloud environments.

Why it matters

By moving these protocols to the Linux Foundation, the industry avoids vendor lock-in, allowing developers to build workflows on specifications that are not controlled by a single entity. This governance model mirrors the success of the CNCF with Kubernetes, providing a stable foundation for cross-company contributions in the rapidly evolving AI agent landscape.