CVE-2025-4143 is a vulnerability classified as CWE-601 (URL Redirection to Untrusted Site, commonly known as an 'Open Redirect') found in the OAuth implementation within the workers-oauth-provider component of the MCP framework by Cloudflare. The core issue arises because the OAuth server implementation fails to properly validate that the redirect_uri parameter during the OAuth authorization flow is included in the allowed list of redirect URIs registered for the client application. Specifically, while the workers-oauth-provider correctly validates the redirect_uri during the token exchange phase, it neglects to perform this critical validation during the initial authorization request. This omission allows an attacker to craft malicious URLs that redirect victims to attacker-controlled sites after OAuth authorization, potentially enabling credential theft or session hijacking. The exploit scenario requires that the victim has previously authorized an OAuth client on the vulnerable server and that the server's authorization logic is configured to auto-approve re-authorizations from previously authorized clients without explicit user interaction. Although the workers-oauth-provider library does not implement this auto-approval logic itself, many applications built on top of it do, which increases the risk. The attacker can lure the victim into visiting a malicious URL that abuses the open redirect flaw to steal OAuth credentials or impersonate the victim on the OAuth server. This vulnerability is a fundamental OAuth security flaw, as the OAuth specification mandates strict validation of redirect URIs to prevent such attacks. The flaw was introduced due to a coding oversight despite awareness of the requirement. The CVSS 4.0 score is 6.0 (medium severity), reflecting network attack vector, high complexity, partial user interaction, and high impact on confidentiality. No known exploits are currently reported in the wild. The vulnerability was published on May 1, 2025, and fixed in a subsequent pull request on the GitHub repository.

For European organizations using the MCP framework's workers-oauth-provider for OAuth authorization, this vulnerability can lead to significant security risks. Attackers exploiting this flaw could steal OAuth credentials, enabling unauthorized access to sensitive resources and services protected by OAuth tokens. This can result in data breaches, unauthorized transactions, and impersonation of legitimate users, undermining trust and compliance with data protection regulations such as GDPR. The impact is particularly severe for organizations relying on automated or streamlined OAuth flows that implement auto-approval of previously authorized clients, as this increases the likelihood of successful exploitation. Additionally, compromised OAuth tokens can be leveraged to pivot within enterprise networks, potentially affecting confidentiality and integrity of critical systems. The medium CVSS score reflects that exploitation requires user interaction and high attack complexity, but the confidentiality impact is high. Given the widespread adoption of OAuth in web services, the vulnerability poses a moderate but tangible threat to European enterprises, especially those in sectors like finance, healthcare, and government that rely heavily on OAuth for secure authentication and authorization.

1. Immediate upgrade to the patched version of workers-oauth-provider that includes the fix for redirect_uri validation during the authorization request phase. Monitor the official GitHub repository for the latest updates. 2. Audit all OAuth server implementations built on workers-oauth-provider to ensure that redirect_uri validation is enforced both during authorization requests and token exchanges. 3. Disable or carefully review any auto-approval logic for OAuth re-authorizations to require explicit user consent, thereby reducing the risk of automated token theft. 4. Implement strict allowlists for redirect URIs and enforce exact matching rather than pattern-based or partial matching. 5. Conduct penetration testing and code reviews focusing on OAuth flows to detect similar misconfigurations or logic flaws. 6. Educate developers and security teams about OAuth best practices, emphasizing the critical nature of redirect URI validation. 7. Deploy web application firewalls (WAFs) with rules to detect and block suspicious OAuth redirect patterns. 8. Monitor OAuth authorization logs for unusual redirect_uri parameters or repeated authorization attempts from the same client. 9. For organizations unable to immediately patch, consider temporarily disabling OAuth clients that rely on workers-oauth-provider or restricting access to trusted networks until remediation is complete.