Despite rapid advances in end-to-end AI music generation, modeling professional Digital Signal Processing (DSP) workflows remains challenging. Existing differentiable plugins often diverge from real-world DSP tools, performing worse than simplified neural controllers given similar computational budgets. We introduce \textbf{WildFX}, a Docker-based pipeline that generates multi-track audio mixing datasets featuring complex effect graphs via a professional Digital Audio Workstation (DAW). \textbf{WildFX} integrates commercial and open-source plugins across VST/VST3/LV2/CLAP formats, supporting advanced signal routing such as sidechains and crossovers, and enabling efficient parallel processing. A streamlined metadata interface simplifies setup. Experiments validate \textbf{WildFX} through blind estimation tasks for mixing graphs and plugin parameters, bridging AI research and practical DSP applications.