Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This repository provides structured workflows for security researchers investigating prompt-to-tool flows, retrieval poisoning, and supply-chain challenges. It focuses on the technical intersection of AI agent orchestration and traditional security boundaries.
The skill set extends to deep infrastructure analysis, including the exploitation of AD CS certificate templates and the bypass of SSL pinning in Android environments.
WHO IT'S FOR
Security researchers
analyzing Android APK hooking and Frida tracing
Penetration testers
exploiting AD CS certificate templates and EKUs
Cloud security engineers
analyzing prompt-injection and supply-chain challenges
CTF competitors
automating security challenges via AI coding clients
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Installs: Complete reverse-skill cybersecurity skills router bundle, including its Kali scripts and the documented skills beyond the bounded skillPaths selection below. · Claude Code, Codex, Cursor, OpenCode, Cline, Windsurf, Kiro
Before you start
Java / JDK
Node.js 22.12+
Python 3.x
A code AI client such as Claude Code, Codex, Cursor, OpenCode, or another compatible client
Platform: Kali Linux
Run from: The cloned reverse-skill repository root—the directory containing README.md
Authorized local binary path replacing <binary-path>
Confirmed case scope
Binary Ninja installation with a valid Vector 35 license
Use the binary-ninja-reverse skill to perform passive triage of the authorized local binary at <binary-path>. Confirm the case scope, verify that Binary Ninja is installed with a valid license, and work from a copy. Report the file hash, architecture, entry points, segments, imports, exports, representative strings, and a few preliminary cross-reference observations. Do not patch the binary or save modifications.
Authorized local project path replacing <project-path>
Confirmed assessment scope
An available SBOM tool recorded in the repository tool index
Use the supply-chain-security skill to review the authorized local project at <project-path>. Check the repository tool index first. If a documented SBOM tool is already available, generate one SBOM and provide a short audit covering unknown or unmaintained dependencies, license conflicts, and direct versus transitive dependencies. If no suitable tool is available, report the missing prerequisite and stop without installing anything, uploading artifacts, or changing the project.
Architecture, configuration, or design-document path replacing <architecture-path>
Confirmed authorization and assessment scope
Use the llm-security skill to create a passive attack-surface map for the authorized LLM application described by <architecture-path>. Identify its LLM-facing entry points, registered tools and parameters, data flow from user input through retrieval and tool calls to outputs, possible system-prompt leakage surfaces, and human-approval gates. Do not send attack payloads or invoke application tools.
Claude
Codex
Cursor
Windsurf
Amp
OpenCode
Kiro
DeepSeek
Cline