Logo
Agent Skills Guide

taint-analysis

published 2026/03/13

Trace attacker-controlled inputs forward to dangerous sinks and backward to discover origins, including deep cross-module taint propagation with parameter mapping, trust boundary analysis, COM vtable resolution, and RPC boundary detection. Reports what sensitive functions are reached, what guards must be bypassed, how tainted data affects internal logic, and identifies multi-module attack chains. Use when the user asks to taint-trace, find where a parameter goes, check if attacker data reaches a sensitive API, identify what needs to be bypassed, trace across DLL boundaries, find privilege escalation vectors, or asks about exploitation potential of a function's inputs.

📦 Install Skill

Command line install

Install directly with the `skills` CLI.

npx skills add marcosd4h/DeepExtractRuntime/skills/taint-analysis

Local download

Download only this skill directory.

Download taint-analysis only
File Explorer
0 files
No file selected
readonly

Select a file to view its content