EEG-PRIME is a novel EEG foundation model for cross-dataset and multi-task decoding. It combines masked pretraining with prototype-aligned instruction tuning to enable instruction-aware and subject-invariant decoding. During pretraining, an EEG encoder learns transferable representations through masked reconstruction with frequency-cutoff spectral augmentation. During instruction tuning, EEG-PRIME incorporates task-semantic, dataset-specific, and subject-invariant conditioning. The resulting conditioning signal modulates the Q-Former through Layer-wise Query Modulation, while frozen text embeddings of class labels serve as prototypes for cosine-similarity-based prediction. Experiments on sixteen datasets show consistent improvements over state-of-the-art baselines and prior EEG foundation models. Blogger's Review: EEG-PRIME is an innovative EEG decoding model with prototype-aligned representation learning and multi-level conditioning. It excels in cross-dataset and multi-task decoding, making it a significant advancement in the field of brain-computer interfaces.