What problem does it solve? Manual code audits often drown in SAST false positives and miss exploitable flaws hidden across files, languages, and supply-chain dependencies. This Skill provides a systematic methodology that prioritizes real attack paths over theoretical defects. ## Core Features & Use Cases - Attack-Surface-Driven Auditing: Enumerate entry points (HTTP routes, message queues, callbacks), trace Source-to-Sink dataflows, and rank findings with a P0-P3 priority matrix. - Multi-Language Vulnerability Patterns: Covers Java, Python, Go, Node.js, PHP, and C# with concrete dangerous-code examples for SQL injection, deserialization, SSRF, SSTI, path traversal, and prototype pollution. - Supply Chain & AI-Assisted Review: Detects dependency confusion, typosquatting, malicious install hooks, and integrates LLM prompt templates with Semgrep/CodeQL pipelines for false-positive filtering and adversarial verification. - Use Case: Given a Java Spring repository, enumerate all unauthenticated endpoints, trace user input to a MyBatis ${} sink, confirm exploitability with a PoC, and produce a CVSS-scored report with a full evidence chain. ## Quick Start Audit this repository for exploitable vulnerabilities by enumerating entry points, tracing dataflows to dangerous sinks, and reporting findings with file and line evidence.
