/

CVE-2026-15481

Severity:

HIGH RISK

(8.8)

OS Command Injection (CWE-77) in Trendnet TEW-635BRM

CVE-2026-15481

Deferred

Vulnerability type

Category:

OS Command Injection

CWE:

CWE-74, CWE-77

Impact:

Remote Code Execution on Network Device

Severity:

HIGH RISK

(8.8)

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

An OS command injection vulnerability in the IPoA WAN Connection Setup of Trendnet TEW-635BRM allows attackers to inject arbitrary shell commands via the ipoa_ipaddr parameter.

When this parameter is processed by the ipoa_test function in /sbin/rc, untrusted input is concatenated into system commands, enabling remote attackers to execute code on affected devices.

Affected context

The issue arises in Trendnet TEW-635BRM devices up to firmware 1.00.03, specifically within the IPoA (IP over ATM) WAN Connection Setup feature.

The vulnerable logic resides in the ipoa_test function of /sbin/rc, where user-provided ipoa_ipaddr values are used to build shell commands.

Deployments where the device's management or configuration interface is exposed to untrusted networks, or where attacker-controlled input can reach the IPoA setup, are at elevated risk.

The devices are end-of-life and no longer receive security support, which increases the long-term exposure if they remain in service.

Exploit preconditions

  • Attack surface: Network-facing management or configuration endpoints that handle IPoA WAN configuration.
  • Required access: Ability to send crafted requests that control the ipoa_ipaddr argument (often via web UI or configuration APIs, depending on deployment).
  • Trigger point: Invocation of the ipoa_test function in /sbin/rc, which constructs and executes shell commands using the supplied parameter.
  • Why exploitation is possible: The application directly interpolates unvalidated ipoa_ipaddr data into command-line strings passed to the system shell, without proper sanitization, escaping, or parameterization, allowing injected tokens to be interpreted as additional commands.

Why this happens

This vulnerability stems from constructing shell commands from untrusted input within firmware scripts.

The ipoa_test function appears to take the ipoa_ipaddr value and concatenate it into a command that is then executed by a shell, rather than treating the value as pure data.

This pattern corresponds to CWE-77: Improper Neutralization of Special Elements used in a Command ('Command Injection'), and is closely related to CWE-74: Improper Neutralization of Data within Output Used by a Downstream Component, because data intended as an IP address is forwarded to the shell without neutralizing metacharacters.

Conceptually, the bug occurs when application logic assumes that configuration parameters like IP addresses are inherently safe, and fails to enforce strict syntax validation or to use safer execution primitives that separate command and arguments.

Potential impact

  • Remote attackers may execute arbitrary OS-level commands with the privileges of the /sbin/rc process.
  • Compromise of the router, including full control of configuration and traffic handling.
  • Installation of persistent malware or backdoors on the device.
  • Use of the compromised router as a pivot point for attacks against internal network hosts.
  • Manipulation, interception, or redirection of user traffic passing through the device.

General mitigation guidance

Mitigation strategies should focus on eliminating direct shell command construction and tightly constraining configuration inputs:

  • Replace shell-based command execution with APIs or system calls that pass arguments as structured parameters, avoiding shell parsing entirely.
  • Apply strict input validation for fields such as IP addresses, ensuring they conform to expected formats and rejecting any characters not needed (e.g., limiting to digits and dots for IPv4).
  • Introduce centralized utilities for safely invoking system-level operations that normalize and validate arguments before execution.
  • Run configuration-handling components with the least privileges required and disable unnecessary services to limit blast radius if a command is injected.
  • For end-of-life hardware, plan for device replacement with supported models that receive security updates and provide more robust management interfaces.

How teams detect issues like this

Detection can rely on both code-centric and operational approaches:

  • Static analysis and firmware review: Scan shell scripts, C code, or embedded web applications for patterns where user input is concatenated into strings passed to system, popen, backticks, exec with a shell, or direct /bin/sh invocation.
  • Configuration-specific checks: During code review, flag any IP address, hostname, or user-supplied network parameter that is reused as part of a command, especially in router and appliance firmware.
  • Policy enforcement: Establish secure coding policies forbidding unsanitized shell invocation, and require security reviews for modules that interface with the operating system.
  • Runtime monitoring: On devices where it is practical, monitor for anomalous process executions or unusual command lines initiated by management processes, which can indicate attempted or successful command injection.

Sources

SourceLink
github.comgithub.com/glkfc/IoT-Vulnerability/blob/main/TRENDnet/TRENDnet_TEW635BRM_ipoa_cmdinj_EN.md
vuldb.comvuldb.com/cve/CVE-2026-15481
vuldb.comvuldb.com/submit/838237
vuldb.comvuldb.com/vuln/377788
vuldb.comvuldb.com/vuln/377788/cti

Frequently asked

questions

Exploitability depends on how the device is deployed.

If the IPoA configuration interface is reachable from untrusted networks or exposed to the internet, exploitation can be fully remote.

In more locked-down environments, attackers would first need a foothold that gives them access to the management plane.

An attacker needs the ability to send crafted input that controls the ipoa_ipaddr value used by the IPoA WAN Connection Setup.

This may require access to the device's administrative interface, or to any upstream component that forwards configuration values to /sbin/rc.

The safest pattern is to avoid constructing shell commands from user input altogether.

Use system APIs that separate code from data, enforce strict validation of configuration parameters, and centralize sensitive operations in well-reviewed helper functions that do not rely on shell parsing.

Embed security into the design of management and configuration paths by:

  • Treating all external input as untrusted, including configuration parameters.
  • Applying whitelisting validators for structured data such as IP addresses and interface names.
  • Prohibiting direct shell invocation in coding guidelines, and using safer execution primitives.
  • Integrating automated security testing and code reviews that specifically search for injection patterns before firmware releases.
Find out what's already
exploitable in your codebase.