Chagas disease is a parasitic illness spread by triatomine bugs that can cause serious heart problems. Mass serological testing is costly, so automated early detection using ECGs is desirable as it is scalable and cost-effective. As part of the George B. Moody PhysioNet Challenge 2025, our team (Chagas_detector) utilized an ECG foundational model (ECG-FM) pretrained on 1.5 million ECGs, and finetuned only an added classification head, leveraging the model's existing feature extraction capabilities, to detect Chagas disease from 12-lead ECGs. The pipeline included signal preprocessing by re-sampling signals to 500 Hz, applying a 0.5 Hz high-pass butterworth filter, followed by powerline filtering, which was then passed into a frozen ECG-FM that encodes the signal into a 768-dimension vector. A classification MLP head, which combines this vector with patient age and sex, was then finetuned to output a binary probability for Chagas disease. We train on all of the PTB-XL, SaMi-Trop, and CODE-15% records. Our model received a Challenge score of .323 (ranked ?) on the hidden validation set.