Anthropic donates the Agent Skills open standard to the Agentic AI Foundation and says 70+ platforms have integrated it, after donating MCP to the AAIF in 2025

More than 70 agent platforms now offer standardized, interoperable skills, allowing you to add specialized capabilities to your AI agents without custom integration.

Anthropic has donated its open standard for AI agent skills, known as Agent Skills, to the Agentic AI Foundation (AAIF). This standard, which allows AI agents to access specialized knowledge and workflows, has already been integrated by over 70 platforms, including prominent ones like ChatGPT, Cursor, GitHub Copilot, and VS Code. The Agent Skills standard was initially published in December 2025 and has since seen over one million skills published and nearly 280 million installed. The donation, announced on October 22, 2026, aims to ensure the standard remains community-driven and vendor-neutral, with the AAIF, co-founded by Anthropic, Block, and OpenAI, providing a neutral governance structure. Anthropic previously donated the Model Context Protocol (MCP) to the AAIF in 2025. Engineers from various companies, including Anthropic, Google, and OpenAI, will continue to maintain the standard as individuals.

The donation of Agent Skills to the AAIF signals a broader trend of major AI players contributing core technologies to neutral foundations to foster ecosystem growth and interoperability. By standardizing how AI agents acquire and utilize specialized knowledge, Anthropic and the AAIF are working to overcome the fragmentation that has plagued AI agent development, making it easier for developers to build and deploy agents that can perform a wider range of complex tasks across different platforms. This move is crucial for maturing the agentic AI landscape from isolated tools to a more connected and functional ecosystem.

The widespread integration of Agent Skills by 70+ platforms, including major developer tools and chatbots, indicates a strong market appetite for standardized agent capabilities. This adoption suggests that the perceived value of interoperability and reusable skills outweighs proprietary solutions, potentially accelerating the development of more sophisticated and versatile AI agents. Future developments will likely focus on versioning, security, and a formal proposal process, as highlighted by the AAIF's proposed roadmap, which will be essential for scaling the standard and ensuring its long-term viability and widespread trust.