CyberStrike

•
3.7k
•

Augments offensive security operations with an AI-powered autonomous red team, automating penetration testing, vulnerability discovery, and exploitation across diverse environments.

CyberStrike is an open-source, AI-augmented offensive security harness designed to transform any large language model into a specialized red team agent. It provides an intelligence layer that injects domain-specific security context, methodologies, and tool orchestration logic into AI interactions, enabling automated reconnaissance, vulnerability discovery, exploitation, and reporting. With support for over 150 AI providers and thousands of models, 13 specialized security agents, 7,600+ Ed25519-signed attack skills, and robust post-exploitation capabilities across various platforms, CyberStrike offers a comprehensive and flexible solution for advanced penetration testing, complete with a web UI and remote tool execution.

Key Features

01Remote tool execution and orchestration across multiple servers with Bolt.
02Autonomous AI agents specializing in web, mobile, cloud, and internal network security.
031,227 GitHub stars
04Over 7,600 Ed25519-signed security skills for deep domain knowledge and attack chaining.
05Built-in HackBrowser for automated and manual HTTP traffic capture and analysis.
06Integration with 150+ AI providers and 5,300+ LLM models for flexible intelligence.

Use Cases

01Vulnerability discovery and exploitation in Linux, Windows, macOS, AWS, Azure, and Kubernetes.
02Automated penetration testing and red teaming for web, mobile, cloud, and network environments.
03Compliance validation against frameworks like MITRE ATT&CK, OWASP WSTG, and CIS Benchmarks.