Anthropic Cybersecurity Skills
Open-source library of 817 structured cybersecurity skills for AI coding agents and security workflows.
🛡️ AgentReady threat assessment
MAESTRO 7-layer threat model + OWASP AIVSS risk score for Anthropic Cybersecurity Skills, derived from its capabilities.
Overview
Anthropic Cybersecurity Skills is an independent, community-created open-source GitHub repository that provides 817 production-grade cybersecurity skills for AI agents. The skills span 29 security domains and follow the agentskills.io open standard. Each skill is mapped to six frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, MITRE D3FEND, NIST AI RMF, and the MITRE Fight Fraud Framework. The repository is designed to be cloned and used with agentic coding and assistant platforms such as Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI, and other compatible platforms. It is not affiliated with Anthropic PBC and is licensed under Apache 2.0.
Key features
- cybersecurity
- security skills
- MITRE ATT&CK
- NIST CSF
- MITRE ATLAS
- D3FEND
- incident response
- threat hunting
- Claude Code
- GitHub Copilot
Use cases
- Adding structured cybersecurity knowledge to coding assistants and autonomous development workflows
- Supporting security investigations with mapped skills for threat hunting, detection, response, and analysis
- Helping security teams align AI-assisted workflows with MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND, NIST AI RMF, and MITRE F3
- Providing reusable cybersecurity skills for Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI, and compatible platforms
- Building security-focused agent workflows without creating a skills library from scratch