diff --git a/.jules/sentinel.md b/.jules/sentinel.md index 0bc6dbd..1b6ff7e 100644 --- a/.jules/sentinel.md +++ b/.jules/sentinel.md @@ -31,3 +31,8 @@ **Vulnerability:** CRLF Injection (CWE-93) / Email Content Injection **Learning:** Interpolating untrusted data (like IP addresses, block reasons, or actor names) directly into email headers or the `Subject` line without sanitization allows an attacker to inject CR/LF characters. This can be used to terminate headers prematurely and inject additional headers (e.g., `Bcc`, `Reply-To`) or even replace the entire email body. **Prevention:** Always sanitize any input destined for a network header or a delimited protocol. Using a `strings.NewReplacer("\r", "", "\n", "")` to strip line-break characters ensures that untrusted content remains confined to its intended field and cannot "break out" to inject new headers. + +## 2026-06-13 - [DOM-based XSS in Toast Notifications] +**Vulnerability:** DOM-based Cross-Site Scripting (XSS) (CWE-79) +**Learning:** Using `innerHTML` with string interpolation to dynamically construct HTML elements using untrusted data allows the browser to execute embedded `