/

CVE-2026-94038

Severity:

HIGH RISK

(7.3)

Server-Side Request Forgery (CWE-918) in dim-sum-app Deno

CVE-2026-94038

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Unauthorized server-side network requests and potential access to internal services

Severity:

HIGH RISK

(7.3)

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 vulnerability in the Deno backend of NonceGeek dim-sum-app allows a remote attacker to influence the supabase_url argument used by textSearchV2Handler.

Under deployment conditions where this value is accepted without strict validation, the server may be induced to send requests to attacker-selected or internal destinations.

Affected context

The vulnerability occurs in the Deno Backend component, specifically the textSearchV2Handler function in deno/main.tsx. It becomes risky when the handler accepts a remotely influenced supabase_url value and uses it to make outbound requests without sufficiently restricting destination schemes, hosts, ports, or network ranges.

Exploit preconditions

  • Attack surface: A remotely reachable request-handling endpoint exposed by the Deno backend
  • Required access: The ability to submit or influence input accepted by textSearchV2Handler
  • Trigger point: Processing of the supabase_url argument during server-side request construction
  • Why exploitation is possible: Destination validation may permit attacker-controlled URLs, allowing the backend to originate requests beyond the intended external service boundary

Why this happens

The underlying bug is insufficient validation and trust-boundary enforcement for a user-influenced URL. When server-side code treats a supplied URL as an authoritative destination, an attacker may redirect backend requests to unintended hosts or network locations.

This behavior maps to CWE-918: Server-Side Request Forgery, in which an application makes outbound requests based on attacker-controlled addressing information.

Potential impact

  • The attacker may access services reachable from the application network
  • Internal administrative or metadata endpoints could be exposed
  • The application could send authenticated requests to unintended destinations
  • Network segmentation boundaries may be weakened
  • Sensitive response data could be disclosed through the vulnerable endpoint
  • Availability may be affected if the server is directed to unsuitable or slow destinations

General mitigation guidance

Mitigation generally focuses on reducing trust in client-supplied destinations and enforcing an explicit outbound request policy.

Common defensive patterns include:

  • Allowlisting approved schemes, hosts, ports, and path patterns
  • Resolving and validating destinations before connecting, including protection against redirects and DNS-based bypasses
  • Rejecting loopback, link-local, private, reserved, and other non-public network ranges unless explicitly required
  • Using a fixed server-side service configuration instead of accepting arbitrary URLs from requests
  • Applying outbound network egress controls and least-privilege service permissions
  • Limiting response handling, request timeouts, and connection targets
  • Avoiding unnecessary credentials on outbound requests

How teams detect issues like this

Organizations commonly detect SSRF risks through:

  • Static analysis rules for HTTP clients that consume request parameters as URLs
  • Code review checks for URL construction, redirect handling, DNS resolution, and IP-range validation
  • Tests covering localhost, private-network, link-local, metadata-service, alternate-scheme, and redirect destinations
  • Runtime egress monitoring for unexpected internal targets or unusual outbound hosts
  • Network policies that restrict backend workloads to approved destinations
  • Automated security review tools that identify attacker-controlled data reaching outbound request APIs

Sources

SourceLink
github.comgithub.com/NonceGeek/dim-sum-app/
github.comgithub.com/NonceGeek/dim-sum-app/commit/8389032e5d52c28c4855c6126ca7d0eae8af346a
github.comgithub.com/NonceGeek/dim-sum-app/issues/366
github.comgithub.com/NonceGeek/dim-sum-app/pull/367
vuldb.comvuldb.com/cve/CVE-2026-94038
vuldb.comvuldb.com/submit/947732
vuldb.comvuldb.com/vuln/407967
vuldb.comvuldb.com/vuln/407967/cti

Frequently asked

questions

Yes. The reported issue is remotely reachable when an attacker can submit input to the exposed handler and influence the supabase_url value.

Authentication requirements depend on the application's endpoint controls. Any user or client able to reach the handler and provide the relevant argument may be able to trigger the vulnerable behavior.

Use a fixed server-side destination or a narrowly defined allowlist, then validate the resolved destination and enforce outbound network restrictions.

Treat all client-provided URLs as untrusted, test URL parsing and redirect edge cases, apply egress policies, and use code review or automated analysis to identify request parameters flowing into network clients.

Find out what's already
exploitable in your codebase.