提出 #992051: SourceCodester Drug Recommendation System Using Machine Learning 1.0 (SourceCodester Release 18278) Improper Authentication (Broken Access Control)情報

タイトルSourceCodester Drug Recommendation System Using Machine Learning 1.0 (SourceCodester Release 18278) Improper Authentication (Broken Access Control)
説明Every protected Admin page uses the same flawed auth guard: it sends a `Location` redirect when `$_SESSION['user_id']` is empty, but then **fails to call `exit()`/`die()`**. In PHP, `header("Location: ...")` merely queues an HTTP header; it does **not** stop execution. As a result the sensitive code that follows the guard — reading datasets, deleting users/patients/symptoms, enabling/disabling accounts, changing passwords, executing SQL — runs even for anonymous, unauthenticated requests. This single class of defect is the linchpin that makes VULN-01 (unauthenticated SQLi), stored XSS rendering, and all destructive operations reachable from outside.
ソース⚠️ https://github.com/hackliu/Vulnerability-Reports/blob/master/Drug-Recommender-Web-App/VULN-02-Broken-Access-Control.md
ユーザー BI1IHA (UID 69245)
送信2026年09月19日 11:27 (1 月 ago)
モデレーション2026年10月05日 20:23 (16 days later)
ステータス承諾済み
VulDBエントリ413697 [SourceCodester Drug Recommendation System 1.0 Auth Guard user_id 弱い認証]
ポイント20

Do you know our Splunk app?

Download it now for free!