UTC
Medium

Portkey.ai Gateway: Server-Side Request Forgery (SSRF) in Custom Host

CVE-2025-66405 · GHSA-hhh5-2cvx-vmfp · published 2025-12-02

Severity
Medium
CVSS score
6.9
Chance of exploitation (EPSS)
0.40%
Fixed in
1.14.0
  • Exploitation No known exploitation
  • Automatable No
  • Technical impact Partial
  • CISA SSVC
What it is

This vulnerability can be exploited to force the server to make requests to arbitrary hosts on the internal network. This could allow an attacker to exfiltrate sensitive data, for instance, by accessing the AWS metadata service.

What to do

Upgrade @portkey-ai/gateway to 1.14.0 or later.

The issue is patched in 1.14.0 (https://github.com/Portkey-AI/gateway/pull/1372) The vulnerability resides within the gateway's request processing function which handles the x-portkey-custom-host header. This parameter was passed directly or with insufficient validation/sanitization to an internal HTTP request function. The fix (v1.14.0) implements a robust allow-list policy: 1. All custom host inputs are now strictly validated to ensure the resulting URI points only to trusted, expected external services. 2. The implementation now explicitly blocks requests to non-routable IP addresses, loopback addresses, private networks and standard metadata endpoints.

Affected
PackageRegistryAffectedFixed in
@portkey-ai/gatewaynpm< 1.14.01.14.0

The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].