Explained briefly
On October 22, Anthropic announced that it is donating Agent Skills to the Agentic AI Foundation, or AAIF, under the Linux Foundation. The format is intended as an open way to give AI agents specialized instructions and workflows.
The verified basic design is unusually simple: a skill is a folder containing a SKILL.md file. That file explains what the AI agent should do and when the knowledge should be used. The folder can also include scripts, reference material, and templates.
Anthropic says more than 70 agent platforms have already integrated the format, including ChatGPT, GitHub Copilot, Cursor, and VS Code. The company also reports that roughly one million skills have been published and installed nearly 280 million times. Those adoption figures come from Anthropic and should be read as company-reported numbers.
Why does this matter? Instead of explaining the same procedure again in every AI service, a recurring job can be packaged as a reusable instruction set. That does not automatically guarantee a good result: quality still depends on the instructions, sources, and any bundled scripts, and an imported skill should be reviewed before it is allowed to run anything.
Our editorial conclusion is that neutral stewardship matters more than the file format itself. If several AI tools can read the same open “recipe,” practical know-how becomes less tied to a single company. Over time, that could make personal routines and organizational workflows both more portable and easier to improve.