Reverse Skill
An open-source AI agent skill pack for reverse engineering and security research, compatible with Claude Code, Codex, Cursor and OpenCode.
Reverse Skill gives AI coding agents workflows for reverse engineering, authorized penetration testing and CTF challenges. It's for security researchers who want their agent to select an analysis method and work with the tools available on their machine.
- Mobile and binary analysis: Workflows cover Android APKs, iOS, native binaries and .NET code, with support for tools such as jadx, apktool, Frida, IDA and Binary Ninja.
- Security investigations: Separate workflows address frontend JavaScript encryption, malware, firmware, HTTP request analysis and penetration testing.
- Task routing: The pack matches a target to a suitable workflow, checks available tools and supports toolchain setup when needed.
- Case review: A dedicated workflow helps review case evidence and prepare reports for handoff, while routing rules require testing scope before target actions.
The pack runs alongside a compatible AI coding client on Windows, Linux or macOS. Claude Code, Codex, Cursor and OpenCode can use the client-independent core. An optional Codex adapter delegates to that core but doesn't register external MCP servers automatically.
The repository is open source under the MIT license. It supplies workflows for an existing coding agent rather than a standalone AI application.
