The vulnerability, described as 'overlong UTF-8 decoding', is located in the minimal UTF-8 decoder within the protobufjs library, specifically in the @protobufjs/utf8 package. Analysis of the security patch 75392ea1b78bdc4faba027b5db44ad7c50e9c454 reveals that the core of the vulnerability lies in the utf8.read function in lib/utf8/index.js. Before the patch, this function would decode non-canonical, overlong UTF-8 byte sequences into characters. This could lead to bypasses of security filters that inspect raw byte data before it is converted to a string. For example, a byte sequence that does not contain an ASCII slash / could be decoded into a string that does. The patch rectifies this by adding validation to detect and reject these overlong sequences, replacing them with the standard Unicode replacement character \ufffd. The new test cases added in lib/utf8/tests/index.js explicitly check for this corrected behavior with known overlong sequences, confirming that utf8.read is the affected function.