A Server-Side Request Forgery (SSRF) vulnerability has been identified in the Angular SSR request handling pipeline. The vulnerability exists because Angular’s internal URL reconstruction logic directly trusts and consumes user-controlled HTTP headers specifically the Host and X-Forwarded-* family to determine the application's base origin without any validation of the destination domain.
Specifically, the framework didn't have checks for the following:
- Host Domain: The
Host and X-Forwarded-Host headers were not checked to belong to a trusted origin. This allows an attacker to redefine the "base" of the application to an arbitrary external domain.
- Path & Character Sanitization: The
X-Forwarded-Host header was not checked for path segments or special characters, allowing manipulation of the base path for all resolved relative URLs.
- Port Validation: The
X-Forwarded-Port header was not verified as numeric, leading to malformed URI construction or injection attacks.
This vulnerability manifests in two primary ways:
- Implicit Relative URL Resolution: Angular's
HttpClient resolves relative URLs against this unvalidated and potentially malformed base origin. An attacker can "steer" these requests to an external server or internal service.
- Explicit Manual Construction: Developers injecting the
REQUEST object to manually construct URLs (for fetch or third-party SDKs) directly inherit these unsanitized values. By accessing the Host / X-Forwarded-* headers, the application logic may perform requests to attacker-controlled destinations or malformed endpoints.
Impact
When successfully exploited, this vulnerability allows for arbitrary internal request steering. This can lead to:
- Credential Exfiltration: Stealing sensitive
Authorization headers or session cookies by redirecting them to an attacker's server.
- Internal Network Probing: Accessing and transmitting data from internal services, databases, or cloud metadata endpoints (e.g.,
169.254.169.254) not exposed to the public internet.
- Confidentiality Breach: Accessing sensitive information processed within the application's server-side context.
Attack Preconditions
- The victim application must use Angular SSR (Server-Side Rendering).
- The application must perform
HttpClient requests using relative URLs OR manually construct URLs using the unvalidated Host / X-Forwarded-* headers using the REQUEST object.
- Direct Header Access: The application server is reachable by an attacker who can influence these headers without strict validation from a front-facing proxy.
- Lack of Upstream Validation: The infrastructure (Cloud, CDN, or Load Balancer) does not sanitize or validate incoming headers.
Patches
- 21.2.0-rc.1
- 21.1.5
- 20.3.17
- 19.2.21
Workarounds
- Use Absolute URLs: Avoid using
req.headers for URL construction. Instead, use trusted variables for your base API paths.
- Implement Strict Header Validation (Middleware): If you cannot upgrade immediately, implement a middleware in your
server.ts to enforce numeric ports and validated hostnames.
const ALLOWED_HOSTS = new Set(['your-domain.com']);
app.use((req, res, next) => {
const hostHeader = (req.headers['x-forwarded-host'] ?? req.headers['host'])?.toString();
const portHeader = req.headers['x-forwarded-port']?.toString();
if (hostHeader) {
const hostname = hostHeader.split(':')[0];
// Reject if hostname contains path separators or is not in allowlist
if (/^[a-z0-9.:-]+$/i.test(hostname) ||
(!ALLOWED_HOSTS.has(hostname) && hostname !== 'localhost')) {
return res.status(400).send('Invalid Hostname');
}
}
// Ensure port is strictly numeric if provided
if (portHeader && !/^\d+$/.test(portHeader)) {
return res.status(400).send('Invalid Port');
}
next();
});
References
A Server-Side Request Forgery (SSRF) vulnerability has been identified in the Angular SSR request handling pipeline. The vulnerability exists because Angular’s internal URL reconstruction logic directly trusts and consumes user-controlled HTTP headers specifically the Host and
X-Forwarded-*family to determine the application's base origin without any validation of the destination domain.Specifically, the framework didn't have checks for the following:
HostandX-Forwarded-Hostheaders were not checked to belong to a trusted origin. This allows an attacker to redefine the "base" of the application to an arbitrary external domain.X-Forwarded-Hostheader was not checked for path segments or special characters, allowing manipulation of the base path for all resolved relative URLs.X-Forwarded-Portheader was not verified as numeric, leading to malformed URI construction or injection attacks.This vulnerability manifests in two primary ways:
HttpClientresolves relative URLs against this unvalidated and potentially malformed base origin. An attacker can "steer" these requests to an external server or internal service.REQUESTobject to manually construct URLs (for fetch or third-party SDKs) directly inherit these unsanitized values. By accessing theHost/X-Forwarded-*headers, the application logic may perform requests to attacker-controlled destinations or malformed endpoints.Impact
When successfully exploited, this vulnerability allows for arbitrary internal request steering. This can lead to:
Authorizationheaders or session cookies by redirecting them to an attacker's server.169.254.169.254) not exposed to the public internet.Attack Preconditions
HttpClientrequests using relative URLs OR manually construct URLs using the unvalidatedHost/X-Forwarded-*headers using theREQUESTobject.Patches
Workarounds
req.headersfor URL construction. Instead, use trusted variables for your base API paths.server.tsto enforce numeric ports and validated hostnames.References