Introduction As agents become integrated with more advanced functionality, such as code generation, you will see more Remote Code Execution (RCE)/Command Injection vulnerabilities in LLM applications.