Skills an agent can carry into a security team's own vocabulary.

Every skill is mapped to named frameworks, so a finding lands in the language the team already reports in. The counts below are the repository's own published figures.

$ git clone Anthropic-Cybersecurity-Skills

github.com/mukul975/Anthropic-Cybersecurity-Skills

817 structured skills ~~published count~~

Mapped to six frameworksNames as published

MITRE ATT&CK

NIST CSF 2.0

MITRE ATLAS

D3FEND

NIST AI RMF

MITRE F3

lanes carry no per-framework counts the repository publishes six names, not a split

817skills

6frameworks

29security domains

20+platforms

Concept mock, not affiliated with the project. Every number and framework name above is quoted from the repository's own description on GitHub Trending, 18 August 2026. The six lanes are drawn at equal width on purpose: the project publishes which frameworks it maps to, not how the 817 skills divide among them, and inventing that split would be the one lie this page could tell.