SpikingForge is an open‑source Python package for building, training, and deploying spiking neural networks (SNNs) on resource‑constrained devices. It features vectorized LIF neurons, trace‑based STDP for unsupervised learning, and a linear readout trained with SGD—all accessible via a simple YAML‑driven CLI. Trained models can export weights as C headers for direct use on microcontrollers like STM32, enabling low‑power edge AI prototyping. The toolchain runs entirely on CPU and focuses on simplicity and reproducibility