/

CVE-2026-72552

Severity:

HIGH RISK

(7.5)

Server-Side Request Forgery (CWE-918): Dub Metatags Endpoint

CVE-2026-72552

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Unauthorized server-side network access and potential data exfiltration

Severity:

HIGH RISK

(7.5)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

A server-side request forgery vulnerability in Dub allows unauthenticated remote attackers to make the server send HTTP requests to caller-selected internal or external hosts through the metatags edge endpoint. This becomes risky when the endpoint fetches arbitrary URLs without authentication or effective destination controls, potentially exposing internal services and cloud metadata.

Affected context

The issue occurs in Dub's metatags edge endpoint when it accepts a caller-supplied URL and uses the server to retrieve that resource. Deployments may be exposed when the endpoint is reachable without authentication and lacks effective restrictions on private, loopback, link-local, or otherwise sensitive destinations.

Exploit preconditions

  • Attack surface: The publicly reachable metatags edge endpoint
  • Required access: No authentication may be required
  • Trigger point: Processing a caller-supplied URL for server-side HTTP retrieval
  • Why exploitation is possible: The endpoint forwards requests to arbitrary destinations without sufficient destination validation or network egress controls

Why this happens

The vulnerability results from treating an external URL supplied by a caller as an acceptable server-side request target.

This maps to CWE-918: Server-Side Request Forgery, where application-controlled network access can be redirected to internal services, protected infrastructure, or external systems. The insecure pattern is direct URL fetching without robust validation, canonicalization, destination policy enforcement, and protection against redirects or alternate address representations.

Potential impact

  • An attacker may scan internal hosts and services
  • The server could access administrative interfaces that are not publicly reachable
  • Cloud instance metadata or temporary credentials may be exposed
  • Sensitive responses could be returned to or exfiltrated by an attacker
  • Internal network access could support lateral movement or follow-on compromise

General mitigation guidance

Mitigation generally focuses on reducing trust in caller-supplied URLs and limiting the server's network reach.

Common defensive patterns include:

  • Allowing requests only to an explicit, minimal set of approved destinations
  • Validating and canonicalizing schemes, hostnames, ports, and resolved addresses
  • Blocking loopback, private, link-local, multicast, and cloud metadata address ranges
  • Revalidating destinations after DNS resolution and across redirects
  • Applying outbound firewall or proxy policies that restrict unnecessary egress
  • Requiring authentication and authorization where the endpoint does not need to be public
  • Returning only the minimum response data needed by the application
  • Using isolated network identities and least-privilege permissions for the fetching service

How teams detect issues like this

Organizations commonly detect SSRF risks through:

  • Static analysis rules for HTTP clients that receive URLs from request parameters
  • Code review checks for missing scheme, host, port, redirect, and IP-range validation
  • Tests covering loopback, private, link-local, metadata, DNS-rebinding, and redirect scenarios
  • Runtime monitoring of outbound requests from the edge service to unusual destinations
  • Egress proxy and firewall policies that alert on blocked or unexpected internal targets
  • Automated security review tools that flag user-controlled URLs passed to network libraries

Sources

SourceLink
github.comgithub.com/dubinc/dub

Frequently asked

questions

Yes. The described endpoint is remotely reachable, and exploitation may not require authentication when an attacker can submit a caller-controlled URL.

The attacker generally needs network access to the metatags edge endpoint and the ability to influence its URL parameter. No separate account may be required.

Avoid unrestricted server-side URL fetching. Use an explicit destination allowlist, validate resolved addresses and redirects, and enforce outbound network restrictions.

Treat all user-supplied URLs as untrusted, apply centralized SSRF protections, restrict service egress, monitor outbound requests, and test against private, loopback, metadata, and redirect-based destinations.

Find out what's already
exploitable in your codebase.