Plagiarism in code repositories presents significant challenges to academic integrity and intellectual property rights. This paper presents a hybrid plagiarism detection system that integrates Term Frequency-Inverse Document Frequency (TF-IDF) analysis with fuzzy string matching based on Levenshtein distance to identify potential code plagiarism. The system leverages the GitHub API to access and analyze code samples from public repositories. Our approach demonstrates improved detection accuracy compared to single-method approaches by combining syntactic similarity metrics with semantic analysis. Experimental results show that this integrated approach effectively identifies both direct code copying and structurally modified plagiarism attempts while minimizing false positives.