Three cybersecurity portfolio projects — fuzzing, agent sandboxing, YARA scanning

three cybersecurity projects that will actually get you hired. This project fuzzes C libraries, which means it feeds them millions of malformed files and then watches for crashes. And those crashes are memory corruption bugs. Memory corruption bugs is how software actually gets exploited. This project assumes AI will get tricked. So instead of filtering what goes in, it controls what the AI is allowed to do. Because prompt injection cannot actually be filtered reliably. The damage has to be contained instead. This project scans a binary against the same kind of detection rules that real malware analysis write, then tells you which rule matched and why it matters. It also does this statically so you can actually analyze the file safely. And just comment projects and I'll send you the source code with the walkthroughs.