/

CVE-2026-40905

Severity:

HIGH RISK

(8.1)

Open Redirect / Host Header Poisoning in LinkAce

CVE-2026-40905

Deferred

Vulnerability type

Category:

Open Redirect / Host Header Poisoning

CWE:

CWE-601

Impact:

Account Takeover via Password Reset Token Theft

Severity:

HIGH RISK

(8.1)

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

LinkAce's password reset feature trusted user-controlled HTTP headers when generating reset URLs, allowing attackers to inject their own domain into the link sent by email.

If a victim clicks a poisoned reset link, the password reset token can be leaked to the attacker's server, enabling full account takeover.

Affected context

This vulnerability affects LinkAce instances that:

  • Are deployed behind proxies or load balancers that set X-Forwarded-* headers
  • Construct absolute URLs for password reset links based on inbound HTTP headers
  • Do not enforce a strict, server-side canonical application host when generating emails

In such setups, user-controlled or proxy-injected headers can directly influence the host portion of password reset URLs.

Exploit preconditions

  • Attack surface: Public-facing password reset endpoint of the LinkAce application.
  • Required access: Ability to trigger a password reset for a target email address (often unauthenticated and public).
  • Manipulable input: Control over X-Forwarded-Host (or equivalent) header in the password reset HTTP request.
  • Application behavior: The application reflects the header value into the absolute URL used in password reset emails without strict validation.
  • Victim action: The victim must click the received reset link, causing their browser to contact the attacker-controlled domain and expose the token.

Why this happens

The core issue is improper trust in client-supplied or proxy-propagated host headers when building security-critical URLs.

Instead of relying on a server-side canonical base URL, the application uses X-Forwarded-Host directly to construct password reset links.

This maps to CWE-601: Open Redirect, extended here to host header poisoning where attacker-controlled host values are embedded into redirects or links.

The abstract pattern is: unvalidated, user-influenced header values are used as part of a security-sensitive URL, allowing redirection or token leakage to attacker domains.

Potential impact

  • Full account takeover by capturing password reset tokens.
  • Unauthorized access to user data stored within the compromised LinkAce account.
  • Lateral movement if compromised accounts share passwords with other services.
  • Phishing and social engineering campaigns leveraging legitimate-looking reset emails.
  • Erosion of user trust in system-generated security emails.

General mitigation guidance

Effective mitigation focuses on eliminating trust in user-controlled host information for security-sensitive URLs:

  • Define and use a canonical application base URL from server-side configuration or environment, never from HTTP headers.
  • When reverse proxies are required, whitelist and validate any forwarded host or proto headers against expected values before use.
  • Treat Host, X-Forwarded-Host, and similar headers as untrusted input; do not embed them directly into application-generated links.
  • For password reset and other tokenized URLs, always:
  • Generate absolute URLs from trusted configuration.
  • Ensure tokens are single-use, short-lived, and bound to the correct application context.
  • Apply defense-in-depth by:
  • Enforcing HTTPS and HSTS to make downgrade or interception harder.
  • Logging anomalous host header usage for review.

How teams detect issues like this

Detection typically combines code-centric analysis with behavior monitoring:

  • Static analysis / code scanning:
  • Search for usages of Host, X-Forwarded-Host, or similar headers in URL construction code.
  • Flag patterns where framework helpers infer full URLs from request objects for security emails or redirects.
  • Code review practices:
  • Review password reset and email generation flows to confirm they use a configured base URL, not request headers.
  • Pay attention to middleware that normalizes proxy headers and how downstream code trusts those values.
  • Runtime monitoring and logging:
  • Log and alert on unusual or unrecognized host values hitting password reset endpoints.
  • Correlate host header anomalies with email sending events for potential poisoning attempts.
  • Automated SAST or security code review tools can be configured with rules to highlight any header-derived host usage in link generation.

Sources

SourceLink
github.comgithub.com/Kovah/LinkAce/security/advisories/GHSA-48wv-jpf4-vjfv

Frequently asked

questions

Yes. An attacker can remotely exploit the issue by sending a crafted password reset request with a manipulated X-Forwarded-Host header, provided they can cause the reset email to be sent to a victim.

Typically no. Password reset endpoints are usually exposed to unauthenticated users. The primary requirement is the ability to initiate a reset for a known email address and control specific HTTP headers.

The most robust fix is to stop deriving security-sensitive URLs from request headers entirely and instead use a server-side canonical base URL defined in configuration. Any header-derived host values should be ignored or strictly validated against this canonical value.

Adopt a security design rule that all externally visible URLs in authentication and account recovery flows are generated from trusted configuration, not from client input. Combine this with secure code review practices that explicitly check URL generation logic and with automated scans for host/header usage in critical flows.

Find out what's already
exploitable in your codebase.