/

CVE-2026-33655

Severity:

HIGH RISK

(7.7)

SSRF via Hostname IP Bypass (CWE-918) in New API

CVE-2026-33655

Analyzed

Vulnerability type

Category:

Server-Side Request Forgery (SSRF)

CWE:

CWE-918

Impact:

Internal Network Access via SSRF

Severity:

HIGH RISK

(7.7)

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

A server-side request forgery (SSRF) flaw in New API's default configuration allows authenticated users to register webhook-like URLs that resolve to internal or metadata IP addresses.

Because hostname-based URLs were not subjected to IP filtering, domain rules could be bypassed to make the system initiate requests into internal networks.

Affected context

This vulnerability appears in New API deployments that:

  • Rely on the default SSRF protection configuration
  • Allow authenticated users to configure notification endpoints (Webhook, Bark, Gotify, etc.)
  • Use hostname-based URLs rather than literal IP addresses

In this context, hostnames are validated only against allow/block lists without resolving and re-checking the resulting IP address, which weakens network egress controls.

Exploit preconditions

  • Attack surface: Application-level notification URL configuration (e.g., Webhook, Bark, Gotify).
  • Required access: Authenticated access with permission to set or modify outbound notification URLs.
  • Trigger point: The system sending a notification to the user-provided URL, causing an outbound HTTP(S) request.
  • Why exploitation is possible: The SSRF filter examines domain names but does not resolve hostnames and enforce IP-based restrictions, allowing crafted hostnames that resolve to internal or metadata IP ranges to pass validation.

Why this happens

The underlying issue is incomplete server-side request forgery protection, aligned with CWE-918: Server-Side Request Forgery (SSRF).

The security control only evaluated hostname-based allow/block rules and did not perform IP resolution and subsequent IP range filtering.

As a result, user-controlled URLs could reference domains that resolve to sensitive internal or metadata IP addresses, leading the gateway to make unintended outbound requests on behalf of the attacker.

Potential impact

  • Gain of indirect access to internal HTTP services that are not exposed externally.
  • Interaction with cloud provider metadata services, which could lead to credential or token exposure.
  • Discovery of internal network structure and service behavior through error messages and timing.
  • Use of the application as a pivot point for further attacks against internal assets.

General mitigation guidance

Mitigation typically focuses on strengthening outbound request controls and reducing trust in user-supplied URLs.

Common defensive patterns include:

  • Resolving hostnames server-side and enforcing strict IP-based allowlists, including re-checking after DNS changes.
  • Denying or tightly restricting access to private, link-local, loopback, and metadata address ranges by default.
  • Segregating network egress for notification components so they cannot directly reach sensitive internal services.
  • Minimizing the set of user-configurable destinations, favoring predefined, vetted endpoints where practical.
  • Applying least-privilege to any credentials used for outbound calls, reducing the impact of SSRF if it occurs.

How teams detect issues like this

Detection efforts can focus on both code-level and operational signals:

  • Static analysis rules that identify user-controlled URLs being passed into HTTP clients without robust SSRF filtering.
  • Code review checklists that explicitly require hostname resolution and IP range validation for outbound connections.
  • Centralized egress logging and anomaly detection to flag unexpected calls to internal or metadata address ranges.
  • Security tests or red-team exercises that attempt to register notification URLs pointing to internal or reserved IP ranges.

Automated code scanning and policy enforcement in CI/CD can help ensure future changes preserve strong SSRF defenses.

Sources

SourceLink
github.comgithub.com/QuantumNous/new-api/commit/20399d3c8fcb4e3649d53163eb11940fd6763743
github.comgithub.com/QuantumNous/new-api/releases/tag/v0.12.0-alpha.1
github.comgithub.com/QuantumNous/new-api/security/advisories/GHSA-6qcr-qxgr-m7fv

Frequently asked

questions

It requires authenticated access to the application with permissions to configure notification URLs, so it is not a fully unauthenticated remote issue but can be abused by compromised or malicious user accounts.

The attacker must be able to log in and set or modify Webhook, Bark, Gotify, or similar outbound notification URLs in the system's configuration.

Implement robust SSRF protections by resolving hostnames, validating the resulting IP against strict allowlists, and blocking private, loopback, and metadata address ranges by default, combined with network-layer egress controls.

Treat all user-supplied URLs as untrusted, mandate SSRF review for any new outbound HTTP feature, enforce IP-based filtering after DNS resolution, and integrate automated checks in code review and testing pipelines to catch missing SSRF controls early.

Find out what's already
exploitable in your codebase.