A high-severity SSRF vulnerability in Google mcp-toolbox enables unauthenticated attackers to force the server to perform unauthorized outbound requests, risking internal network exposure.

The generic HTTP source and tool components lack sufficient validation of user-supplied URLs, allowing for Server-Side Request Forgery (SSRF). This vulnerability is accessible to unauthenticated attackers.

An attacker can leverage this SSRF to scan internal network infrastructure, interact with internal APIs, or access restricted cloud metadata services. The CVSS score of 8.0 underscores the risk of lateral movement and potential exposure of sensitive internal resources.

Immediate Action: Apply the vendor-provided security update immediately. The fix is associated with commit d67cfbe8ddca5a50c9577ea11feaf5c94d0acace in the upstream repository.

Proactive Monitoring: Inspect outbound network traffic logs from the server for unexpected connections to internal IP addresses or sensitive cloud service endpoints.

Compensating Controls: Enforce strict egress filtering on the host running mcp-toolbox to prevent it from initiating unauthorized connections to internal resources.

Public Exploit Available: No

SSRF vulnerabilities can bypass perimeter security, making them highly dangerous. Administrators should prioritize patching this issue to secure the server against internal network reconnaissance and potential data theft.