/

CVE-2026-22681

Severity:

HIGH RISK

(8.5)

Server-Side Request Forgery (CWE-918) in OpenViking

CVE-2026-22681

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Unauthorized access to internal network services and sensitive resources

Severity:

HIGH RISK

(8.5)

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

OpenViking contains a server-side request forgery vulnerability in its resources API that allows authenticated, low-privilege users to cause the server to make requests to arbitrary URLs.

Because redirects are followed and responses can be retrieved through normal content APIs, attackers may access internal services, cloud metadata endpoints, or other protected network resources.

Affected context

The vulnerability occurs in OpenViking deployments that expose the resources API and permit authenticated users to submit URLs for server-side retrieval. Risk is higher when the application can reach loopback interfaces, RFC 1918 networks, link-local addresses, cloud metadata services, or internal administrative endpoints, and when retrieved responses are exposed through content APIs.

Exploit preconditions

  • Attack surface: The resources API endpoint and related content-retrieval APIs
  • Required access: An authenticated account with low-level privileges sufficient to submit resource URLs
  • Trigger point: Server-side processing of a user-supplied URL
  • Network condition: The OpenViking server must be able to reach an internal or otherwise restricted destination
  • Why exploitation is possible: The server follows redirects and returns retrieved content without adequately restricting destination addresses or validating the final request target

Why this happens

The underlying bug is server-side request forgery, mapped to CWE-918. The application treats attacker-controlled URLs as trusted outbound destinations and performs network requests from a privileged server-side context.

Insufficient destination validation, redirect handling, and network egress controls can allow requests to reach loopback, private, link-local, or cloud metadata addresses. Exposing the resulting responses through normal application APIs further enables internal service discovery and interaction.

Potential impact

  • Attackers may enumerate internal hosts and services
  • Attackers could retrieve sensitive cloud instance metadata or service credentials
  • The application may expose administrative interfaces that are not publicly reachable
  • Internal services could receive unintended requests from a trusted network position
  • Sensitive response data may be disclosed through content APIs

General mitigation guidance

Mitigation generally focuses on reducing trust in user-provided URLs and enforcing destination-aware outbound request controls.

Common defensive patterns include:

  • Allowing requests only to explicitly approved schemes, hosts, and ports
  • Resolving hostnames and validating every destination address before connecting
  • Repeating destination checks after redirects or disabling redirects by default
  • Blocking loopback, private, link-local, multicast, and cloud metadata ranges
  • Applying network egress policies that restrict application destinations
  • Avoiding direct exposure of fetched response bodies when they are not required
  • Running the service with minimal network reachability and least privilege
  • Separating resource-fetching components from sensitive internal networks

How teams detect issues like this

Detection approaches can include:

  • Static analysis for HTTP clients that consume user-controlled URLs
  • Code review checks for missing scheme, hostname, port, DNS, and redirect validation
  • Tests covering private, loopback, link-local, and metadata destinations
  • Monitoring outbound requests from application processes to unusual internal addresses
  • Alerting on resource API activity followed by access to internal service ranges
  • Policy enforcement requiring approved egress destinations for server-side fetchers

Automated security review tools can help identify URL-fetching code paths, but runtime network monitoring and targeted integration tests remain important.

Sources

SourceLink
github.comgithub.com/volcengine/OpenViking/commit/41e345896d247e43ab78bbcb38b4a5b1b38ef62c
github.comgithub.com/volcengine/OpenViking/pull/1133
github.comgithub.com/volcengine/OpenViking/releases/tag/v0.3.4
vulncheck.comvulncheck.com/advisories/openviking-ssrf-via-api-v1-resources
github.comgithub.com/volcengine/OpenViking/pull/1133https://github.com/volcengine/OpenViking/pull/1133

Frequently asked

questions

It can be exploitable through the exposed API by an authenticated user, provided the OpenViking server can reach the targeted internal or restricted destination.

No. The described issue requires authentication and a low-privilege account with permission to submit resource URLs.

Treat submitted URLs as untrusted, enforce an allowlist of permitted destinations, validate resolved addresses, and apply the same controls after redirects.

Combine strict URL validation with network egress restrictions, least-privilege deployment, monitoring of outbound requests, and tests for internal-address access.

Find out what's already
exploitable in your codebase.