This rule detects command injection, a serious security vulnerability that occurs when an attacker is able to manipulate a command that the application executes. Command injection attacks can lead to data loss, corruption, or unauthorized access to sensitive system data.
Command injection vulnerabilities generally occur when user input is used unsanitized in a command that is executed by the application. In the non-compliant code samples, the user input is directly added to a command that is executed by the application, allowing an attacker to potentially execute arbitrary commands.
To avoid this vulnerability, user input should never be used directly in a command that is executed by the application. Instead, use safe APIs that allow you to execute commands without the risk of injection, or ensure that user input is properly sanitized before it is used. If you must use user input in a command, ensure that it is properly escaped or quoted to prevent the injection of additional commands.
importjavax.servlet.http.HttpServletRequest;importjavax.servlet.http.HttpServletResponse;publicclassCompliant{publicvoiddoPost(HttpServletRequestrequest,HttpServletResponseresponse){// Safe: using hardcoded command without user inputProcessBuilderpb=newProcessBuilder();pb.command("ls","-la");}}
원활한 통합. Datadog Code Security를 경험해 보세요
Datadog Code Security
이 규칙을 사용해 Datadog Code Security로 코드를 분석하세요
규칙 사용 방법
1
2
rulesets:- java-security # Rules to enforce Java security.
리포지토리 루트에 위의 내용을 포함하는 static-analysis.datadog.yml을 만듭니다
무료 IDE 플러그인을 사용하거나 CI 파이프라인에 Code Security 검사를 추가합니다