Summary
Internet Routing Registry daemon (IRRD) v4 suffers from a combination of Host header poisoning and insecure password reset/account activation link generation.
By manipulating the HTTP Host header, an attacker can cause IRRD to send confirmation emails containing links to an attacker-controlled domain, capture the token when the victim clicks, and then use it on the legitimate IRRD instance to take over accounts that do not have two-factor authentication enabled.
Affected context
This vulnerability appears in IRRD v4 deployments that:
- Expose the web interface for password reset and account creation over HTTP(S)
- Derive absolute URLs for confirmation links from the incoming HTTP Host header
- Use email-based confirmation tokens for password reset or new account activation
- Do not enforce strict validation of the Host header or canonical application URL
In such environments, attackers can abuse crafted requests to influence the confirmation link domain used in outbound emails.
Exploit preconditions
- Attack surface: Public-facing IRRD HTTP(S) interface handling account creation and password reset flows.
- Required access: Ability to send HTTP requests to the IRRD web endpoint (no prior authentication required for initiating password reset/account creation).
- User interaction: A legitimate user (or the new-account recipient) must open the confirmation link in the email.
- Trigger point: The server constructs the confirmation URL from the unvalidated HTTP Host header and sends it via email.
- Why exploitation is possible: The application trusts the Host header as a source of truth for the base URL and does not bind tokens to a specific domain or origin, allowing the attacker to intercept tokens by directing victims to an attacker-controlled domain.
Why this happens
The core problem is insecure construction of externally visible links using untrusted request metadata, specifically the HTTP Host header.
This maps to CWE-601: Open Redirect in the broader sense of untrusted URL generation and CWE-640: Weak Password Recovery Mechanism for Forgotten Password, because the reset mechanism exposes a token in a way that can be intercepted.
IRRD uses the Host header to build absolute URLs in password reset and account creation emails without validating that it matches a canonical, trusted hostname.
When a victim follows the attacker-controlled link, the embedded reset/confirmation token is sent to the attacker's domain first, giving them full access to use the token on the legitimate IRRD instance and complete account takeover for accounts without enforced two-factor authentication.
Potential impact
- Attackers can take over IRRD user accounts that do not have two-factor authentication enabled.
- Malicious modification of RPSL objects maintained by the compromised account's mntners.
- Unauthorized changes to routing registry data, potentially impacting routing policies and network trust relationships.
- Performance of any actions available to the compromised account, including administrative actions if the account has elevated privileges (short of those protected by mandatory 2FA).
- Damage to data integrity and trust in the routing registry ecosystem, with potential downstream operational impact on networks relying on affected data.
General mitigation guidance
Effective mitigation focuses on removing trust in unvalidated request headers and hardening account recovery flows:
- Derive the application's base URL from a trusted configuration value rather than the HTTP Host header or other client-controlled fields.
- Enforce strict validation of the Host header against an allowlist of expected hostnames, rejecting or normalizing unexpected values.
- Bind reset and confirmation tokens to specific contexts (e.g., IP, user agent, or short lifetimes), while recognizing privacy and reliability trade-offs.
- Use opaque, single-use, time-limited tokens for password reset and account activation, and invalidate them immediately after use.
- Encourage and, where appropriate, require multi-factor authentication for sensitive accounts to reduce the impact of token compromise.
- Apply least-privilege principles to user roles so that compromise of a single account has constrained blast radius over routing objects and override capabilities.
How teams detect issues like this
Detection can rely on both application-level reviews and operational monitoring:
- Static analysis and code review:
- Flag usage of request.Host (or framework equivalents) when constructing absolute URLs for emails or redirects.
- Look for functions that build links from user-controlled inputs without cross-checking against a canonical base URL.
- Review password reset and account creation logic for token issuance, scoping, and validation practices.
- Configuration and policy checks:
- Enforce policies that all public URLs are derived from configuration, not from headers.
- Validate that Host header allowlists or reverse-proxy normalization are in place.
- Runtime monitoring:
- Monitor logs for anomalous Host header values, especially on password reset and account creation endpoints.
- Alert on patterns where many reset/account emails are associated with unusual hostnames or sudden spikes in reset requests.
- Automated application security testing tools can be configured to fuzz Host headers on authentication and password recovery endpoints to surface similar weaknesses in other services.

