Anthropic announced on October 22, 2026 that it is donating Agent Skills, an open standard for giving AI agents specialized knowledge and workflows, to the Agentic AI Foundation (AAIF) under the Linux Foundation. Anthropic says it co-founded the AAIF with Block and OpenAI last year and previously donated the Model Context Protocol (MCP) to it. According to Anthropic, a skill is a folder containing a SKILL.md file that teaches an agent a specific task or domain, along with any scripts or reference material it needs. The company says it developed skills while building Claude Code, launched them in October 2025, and published the open standard in December 2025. Anthropic states that more than 70 agent platforms, including ChatGPT, Cursor, GitHub Copilot, and VS Code, have integrated the standard, and that 1 million skills have been published and nearly 280 million installed. The donation covers the Agent Skills name and logo, the GitHub organization, and the agentskills.io domain. Anthropic says engineers from Anthropic, Google, OpenAI, and other companies will maintain the standard as individuals, with David Soria Parra and Mahesh Murag as Lead Maintainers. A proposed roadmap covers versioned releases, a formal proposal process, and security guidance. These figures come from Anthropic's own announcement.
- Agent Skills donated to the Agentic AI Foundation under the Linux Foundation, announced October 22, 2026
- Anthropic says the standard was published in December 2025 and integrated by 70+ platforms including ChatGPT, Cursor, GitHub Copilot, and VS Code
- Anthropic reports 1 million skills published and nearly 280 million installed
- Donation includes the name, logo, GitHub organization, and agentskills.io domain; a proposed roadmap includes security guidance for writing and installing skills
What it means for you
This is mostly plumbing news for people who build AI tools, not a decision point for most operations teams. The practical takeaway: 'skills' are just folders of instructions that teach an AI agent how to do a specific task your way, and the format for writing them is now standardized across the major AI tools your team may already use. Putting it under a neutral foundation means you're less likely to be locked into one vendor's version. The one thing worth flagging: the announcement itself says security guidance is still a proposed, future item, so installed skills are instructions an agent will follow — treat them like code you didn't write.
Try this
If your team already uses a tool like GitHub Copilot, Cursor, or Claude Code, ask whoever owns it to show you one 'skill' it has installed and what instructions are inside it — before deciding whether any are worth writing for your own repeatable tasks.
Who should care
Developers and the people who manage your AI coding or agent tools, plus whoever is responsible for software security review.
Skip this if
Your organization isn't building with AI agents or developer tools yet. This is standards-body news with no action for a general operations team.
Sources: Anthropic — read the original