CVE-2026-53666
Medium · 6.1CVSS
6.1
EPSS
0.4%
Published
Jul 27, 2026
CISA due date
—
Cloudanix analysis
In certain web applications that handle server errors in a specific way, an attacker could make users' browsers contact outside servers, but only with unusual code that most apps do not have and the issue is now fixed by an update.
- Exploitation
- Requires User Interaction
- Attacker needs
- User Interaction
- Complexity
- High
- Impact
- Other
- Component
- Web Framework
- Class
- Improper Input Validation
- Remediation
- Upgrade
- Effort
- Medium
- Patch risk
- Low
What it actually does
Improper input validation in SSR error handling allows attacker-supplied input to overwrite error object properties, triggering unexpected constructor execution during client-side hydration that initiates outbound network requests. Limited to Framework Mode and Data Mode applications performing manual SSR/hydration; Declarative Mode is unaffected.
Why it matters to the business
Low-probability risk of unauthorized client-initiated network callbacks in specific SSR setups, potentially enabling limited data leakage or external callbacks. Affects applications using shopify react-router 6.4.0-7.17.0 with custom error handling during hydration; low EPSS and no known exploitation reduce immediate operational urgency despite 91 days since publication.
How an attack unfolds
An attacker identifies an application using manual SSR/hydration in Framework or Data Mode, supplies crafted input that overwrites aspects of SSR-caught errors, and triggers error paths that propagate to client hydration where constructor execution causes an outbound network request.
How to fix it
Upgrade shopify react-router to version 7.18.0 or later. Review and avoid manual SSR error handling patterns that permit untrusted input to influence error objects.
How to confirm the fix worked
Inspect package.json and installed node_modules for react-router versions prior to 7.18.0; test SSR error scenarios with controlled inputs to verify constructor execution is no longer triggered on hydration.
Typical attack path
- Identify target application using React Router Framework/Data Mode with manual SSR/hydration
- Supply attacker-controlled input that reaches SSR error handling paths
- Overwrite specific error object properties during SSR
- Trigger client hydration that executes attacker-influenced constructor
- Initiate outbound network request from the client
How to detect it
Examine server-side rendering logs for anomalous error inputs and client-side console or network logs for unexpected constructor invocations or hydration-triggered requests; correlate with application error handling paths.
What your security team should do
- Upgrade to react-router 7.18.0 or newer in all affected applications
- Audit custom SSR and error-handling code for untrusted input paths
- Monitor client-side network activity for unexpected requests originating from hydration logic
The record
React Router is a router for React. In versions 6.4.0 through 7.17.0, if application code was written in a way that allows attacker-supplied input to overwrite certain aspects of errors caught by the SSR process, then it was possible for an attacker to trigger unexpected constructor execution on the client, which would in turn trigger an outbound network request. This is only possible with very specific (and unlikely) application-layer code. Note that this does not impact an application if it is using Declarative Mode. It only impacts Framework Mode and Data Mode applications that perform manual SSR/hydration. This issue has been fixed in version 7.18.0.
Affected packages
| Ecosystem | Package | Introduced | Fixed |
|---|
Related vulnerabilities
- CVE-2026-119578 MONAI from 1.6.0 before 1.6.1 contains a remote code execution vulnerability in Auto3DSeg… CWE-470
- CVE-2026-108097 Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection')… CWE-470
- CVE-2026-108099 Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection')… CWE-470
- CVE-2026-119050 Agno through 3.1.11 contains an unsafe reflection vulnerability in PandasTools that… CWE-470
- CVE-2026-116692 MLflow is an open source AI engineering platform for agents, large language models, and… CWE-470
- CVE-2026-115931 Postiz lets a signed-in user ask the backend to run a function on one of their connected… CWE-470
- CVE-2026-115886 MLflow is an open source AI engineering platform for agents, large language models, and… CWE-470
- CVE-2026-115749 hydra-ray-launcher 1.1.0.dev1 before 1.2.2 and 1.4.0.dev4 before 1.4.0rc2 contains an… CWE-470