Summary
Webstudio contains an unauthenticated server-side request forgery vulnerability in proxy routes that accept attacker-controlled URLs when the RESIZE_ORIGIN environment variable is unset. An attacker may use these routes to access cloud instance metadata, reach internal services, and perform network reconnaissance from the server's infrastructure.
Affected context
The vulnerability affects Webstudio deployments through version 0.296.0 when the /cgi/image, /cgi/video, or /cgi/asset proxy routes are exposed and the RESIZE_ORIGIN environment variable is unset. In this configuration, externally supplied URLs may be forwarded by the server without sufficient restriction on the destination.
Exploit preconditions
- Attack surface: Publicly reachable Webstudio image, video, or asset proxy routes
- Required access: No authentication, if the affected routes are exposed without access controls
- Trigger point: A request that causes the proxy to fetch an attacker-selected URL
- Why exploitation is possible: The server may make outbound requests to destinations that are not adequately restricted when the origin-control configuration is absent
Why this happens
The underlying issue is server-side request forgery, mapped to CWE-918. A proxy endpoint accepts or derives a remote URL from request-controlled data and retrieves it from the server environment without enforcing a narrowly defined destination policy. This allows network access to be evaluated from the server's location rather than the attacker's, potentially exposing internal address ranges and metadata services.
Potential impact
- An attacker may read cloud instance metadata and temporary credentials
- An attacker could access internal administrative or application services
- The server may be used for network reconnaissance
- Internal endpoints could be probed or abused through the trusted server network
- Sensitive data may be disclosed from reachable services
General mitigation guidance
Mitigation generally focuses on reducing trust in user-supplied URLs and making outbound proxy behavior deny-by-default.
Common defensive patterns include:
- Restricting proxy destinations to an explicit allowlist of approved origins
- Requiring a configured, validated origin policy before proxy routes are enabled
- Blocking loopback, link-local, private, reserved, and cloud metadata address ranges
- Revalidating DNS results and the final connection destination to reduce DNS-rebinding risks
- Limiting outbound network access with firewall rules, egress policies, or service-mesh controls
- Running the application with only the minimum permissions needed and limiting access to instance metadata
- Applying authentication and authorization where public proxy access is not required
How teams detect issues like this
Organizations can identify SSRF exposure through:
- Static analysis for server-side HTTP clients that consume request-derived URLs
- Code review checks for missing URL allowlists and private-address filtering
- Dynamic testing of proxy routes against loopback, private, link-local, and metadata destinations in controlled environments
- Network monitoring for unexpected egress from Webstudio to internal address ranges
- Policy enforcement requiring explicit outbound destination controls for fetch or proxy features
- Automated security review tools that flag URL-fetching endpoints without validation and egress restrictions

