CyberStrikeAI is a full-stack, AI-native cybersecurity operations platform designed for autonomous penetration testing and red teaming. It orchestrates multiple AI agents using the CloudWeGo Eino ADK framework to plan and execute complex security assessments. The platform manages its own C2 (Command & Control) infrastructure with encrypted beacons and listeners, integrates with over 100 security tools via the Model Context Protocol (MCP), and maintains a persistent knowledge base and project fact graph. It features a comprehensive governance layer including Role-Based Access Control (RBAC), Human-in-the-Loop (HITL) approval workflows, tool execution guards, and detailed audit logging. The architecture is a monolithic Go-like backend serving a web SPA, with plugins for Burp Suite and Chrome. It supports multi-agent orchestration modes (single, deep, plan-execute, supervisor), a graph-based workflow engine, and integrations with various chat platforms (WeChat, Discord, Telegram). The codebase demonstrates high craftsmanship in its execution governance, process isolation, atomic update mechanism, and database schema reconciliation. A significant technical concern is the use of a non-standard Go dialect (featuring `func`, `type`, and `map` literals) tied to the Eino framework. This creates a hard dependency on a specific toolchain and represents a major drift risk from the standard Go ecosystem, limiting portability and the ability to use standard tooling. Despite this, the project is actively maintained with extensive test coverage and a clear focus on production-grade security automation.
Mythic is a cross-platform, post-exploitation, red teaming command-and-control (C2) framework. It enables operators to manage remote agents (callbacks), generate payloads, issue tasks, exfiltrate files, harvest credentials, and automate workflows through an event-driven system. The architecture is a multi-container Docker deployment: a Go backend (using Gin, RabbitMQ, PostgreSQL, gRPC) handles agent communication, task orchestration, and data persistence; a React frontend (Material-UI, Apollo Client) provides a rich operator interface with real-time updates via GraphQL subscriptions (backed by Hasura); and a CLI tool (Mythic_CLI) manages container lifecycle, configuration, and service installation. Key entry points include the agent message webhook (HTTP POST/GET), web controllers for CRUD operations, and RabbitMQ RPC handlers for internal service communication. External dependencies include RabbitMQ, PostgreSQL, Hasura GraphQL engine, Docker, and gRPC. The system is designed for extensibility through payload type containers, C2 profiles, translation containers, and custom browser scripts.
DuplexSpyCS is a full-featured Remote Access Trojan (RAT) command-and-control (C2) framework written in C# (.NET WinForms). It consists of a server-side management console and multiple payload variants (winClient48, winClient48DLL, Tipoff, etc.) that provide comprehensive post-exploitation capabilities: file manager, remote desktop, keylogger, webcam/microphone streaming, registry editor, process/task manager, remote shell (including xterm via WebView2), SOCKS5 proxy, and a plugin system for credential theft (Chrome, Edge, Firefox, MobaXterm). Communication uses a custom binary protocol (clsDSP) over TCP, TLS, or HTTP with RSA-AES hybrid encryption. The payloads employ persistence via registry/startup, UAC bypass, process injection (APC, EarlyBird, CreateRemoteThread, etc.), and fileless execution. The codebase is extensive but exhibits significant code duplication across payload projects, many incomplete or stub features, and outdated dependencies (AForge, NAudio, SharpPcap). It is explicitly designed for offensive security research and learning, but carries high risk of misuse.
BYOB (Build Your Own Botnet) is an open-source post-exploitation framework that provides a command-and-control (C2) server, a custom payload generator, and a suite of post-exploitation modules. It is designed for security researchers and developers to simulate botnet operations for educational purposes. The system consists of a console-based C2 server (Python) and a Flask-based web GUI for managing sessions, payloads, and exfiltrated data. The architecture follows a classic botnet pattern: clients (reverse TCP shells) connect to the C2 server, authenticate via Diffie-Hellman key exchange with AES-256-CBC encryption, and receive tasks (modules) that are remotely loaded and executed. Modules include keylogging, screenshot capture, privilege escalation, packet sniffing, port scanning, persistence mechanisms, and interaction with iCloud/Outlook. The web GUI offers a dashboard, payload generator, terminal interface, and file management. The codebase shows signs of significant duplication (identical modules exist in both the core and web-gui directories) and relies on heavily outdated third-party frontend libraries (CKEditor 2013, FullCalendar 2014, CodeMirror addons). The project appears to be in a stale maintenance state with known incomplete features (e.g., unencrypted remote imports, limited transport types). While the core concepts (remote module loading, encrypted C2, cross-compilation) are technically interesting, the implementation quality is uneven, with Python 2/3 compatibility hacks, security-sensitive patterns (exec usage, SimpleHTTPServer without auth), and a large volume of unused/purely vendored assets inflating the repository. It is best treated as a reference architecture for botnet design rather than production-ready software.