Skip to content
Overseas3 min full readContains AI-generated content · human-reviewed

Anthropic hands its agent-skill format to a neutral foundation

Anthropic donated Agent Skills, its open standard, to the Linux Foundation's Agentic AI Foundation.

  • Platforms & agents
  • Application layer
A metal fastener clamping a stack of paper into one document
Library photography for illustration, not event footage; depicted people and brands are unrelated to the events covered.

01What happened

A skill is a folder holding a SKILL.md file plus any scripts, templates and references, so an agent can load a body of procedural knowledge on demand. Since the standard shipped in December 2025, more than 70 platforms have integrated it, including ChatGPT, Cursor, GitHub Copilot and VS Code.

By the published figures: about 1M skills in the open, with roughly 280M installs. The donation covers the name, the logo, the GitHub organisation and the domain; the specification is licensed Apache 2.0 and CC BY 4.0.

The same day the foundation took on three other projects — a cross-vendor model interface spec, a reinforcement-learning environment framework and a sandbox platform — taking it from 3 hosted projects to 10.

02By the numbers

70+
platforms that have integrated the standard
1M
skills published
280M
cumulative installs

03Background

The same day the foundation took on three other projects, bringing its total to 10. The previous agent primitive given to the foundation was MCP.

Related items

A model training interface open on a laptop screen

Google opens Gemini 4 Argon with a 1M-token output ceiling

On 13 October US Eastern (14 October in Beijing), Google opened Gemini 4 Argon globally: a single-response output ceiling of 1M tokens, a 1,048,576-token context shared across input, thinking and output, available through AI Studio, the Gemini API and Antigravity.

Read the record
A model interface open on a laptop screen

Meta publishes the weights for Muse Spark Open

Meta released Muse Spark Open, an open-weights model that runs on two paired Mac desktops and scores 39 on the Artificial Analysis Intelligence Index. Meta calls it the most capable open-weights model from the US.

Read the record

Want to apply these facts to your business? Bring the question Start AI diagnostic

This section is maintained by CodeLeo: topic and structure are ours, imagery comes from a licensed library, and the text runs through fact sheet → AI-generated draft → human review. The copy therefore contains AI-generated content. It is a public record of facts, states facts only, carries no commentary, and does not provide newsgathering or syndication services.

Company, product and brand names appear only to identify them. This section has no affiliation, agency, sponsorship or partnership with those companies, and nothing here recommends or endorses their business.