Severity:

HIGH RISK

(7.2)

Command Injection in ipTIME C200 ApplyRestore Endpoint

CVE-2026-7833

Deferred

Vulnerability type

Category:

Command Injection

CWE:

CWE-74, CWE-77

Impact:

Remote Code Execution on Embedded Device

Severity:

HIGH RISK

(7.2)

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

EFM ipTIME C200 firmware up to version 1.092 contains a command injection flaw in the ApplyRestore endpoint of /cgi/iux_set.cgi, where the RestoreFile parameter is passed to system-level commands without proper sanitization.

A remote attacker who can reach the device's web interface may inject arbitrary shell commands through this parameter, potentially taking full control of the affected camera.

Affected context

The vulnerability occurs in the web management interface of the ipTIME C200 network camera, specifically in the ApplyRestore functionality exposed via /cgi/iux_set.cgi.

It is relevant in deployments where:

  • The device's HTTP or HTTPS management interface is reachable over the local network or internet
  • Firmware version is 1.092 or earlier
  • The ApplyRestore endpoint is enabled and accessible, and the RestoreFile parameter is processed with insufficient validation

Because this is an embedded device, it may operate unattended on less protected networks, increasing the risk of opportunistic exploitation.

Exploit preconditions

  • Attack surface: HTTP(S) endpoint /cgi/iux_set.cgi on the ipTIME C200 camera
  • Required access: Network-level ability to send HTTP requests to the device; additional authentication may or may not be required depending on configuration
  • Trigger point: Invocation of the ApplyRestore functionality with a crafted RestoreFile argument that includes shell metacharacters or injected commands
  • Why exploitation is possible: User-supplied input from RestoreFile is concatenated into or otherwise incorporated into command-line arguments executed by the system shell, without adequate escaping, validation, or use of safer APIs

Why this happens

The underlying issue is an unsafe construction of operating system commands from web request parameters.

The ApplyRestore handler in /cgi/iux_set.cgi uses the RestoreFile parameter directly in shell commands, likely via functions such as system(), popen(), or equivalent wrappers, without neutralizing shell metacharacters or restricting allowed values.

This maps to:

  • CWE-77: Improper Neutralization of Special Elements used in a Command ('Command Injection'), because untrusted data controls part of a system command
  • CWE-74: Improper Neutralization of Special Elements in Output Used by a Downstream Component, as the HTTP-layer input is forwarded to the OS command layer without appropriate encoding

The abstract pattern is: input from an external interface is treated as data but is embedded into a command string that a downstream interpreter (the shell) executes.

Potential impact

  • Execution of arbitrary shell commands on the camera's operating system
  • Installation of persistent malware or backdoors on the device
  • Use of the camera as a foothold for lateral movement within the local network
  • Unauthorized access to or manipulation of camera video streams and configuration
  • Participation of the device in botnets or distributed denial-of-service (DDoS) attacks
  • Loss of confidentiality and integrity of network credentials stored or proxied by the device

General mitigation guidance

Effective mitigation typically involves eliminating direct command construction from user input and narrowing trust in configuration parameters:

  • Replace shell command concatenation with safer APIs that do not invoke the shell (e.g., direct function calls or exec-style calls with argument arrays and no shell parsing).
  • Constrain the RestoreFile parameter to a strict allowlist of expected values (such as predefined filenames or IDs) rather than arbitrary paths or strings.
  • Apply rigorous input validation and canonicalization, rejecting any value containing shell metacharacters or unexpected characters.
  • Separate concerns between user input handling and system-level operations, ensuring that only validated, normalized data flows into privileged code paths.
  • Run the web interface and restore-processing components with least-privilege accounts and restrict file system access, so that even if a flaw exists the blast radius is reduced.
  • Consider defense-in-depth controls at the network layer (e.g., limiting management interface exposure to trusted segments) to reduce exploitable surface.

How teams detect issues like this

Detection can focus on both code-level patterns and runtime behavior:

  • Static analysis and code review:
  • Flag uses of system(), popen(), exec* with concatenated strings derived from HTTP parameters.
  • Look for string-building routines that combine raw request fields (like RestoreFile) with command-line fragments.
  • Enforce secure coding guidelines that prohibit direct shell invocation from web-layer handlers.
  • Configuration and policy checks:
  • Apply internal security policies that require parameterized, shell-free APIs for system operations.
  • Integrate automated code review tools or SAST scanners into the firmware build pipeline to identify command injection patterns.
  • Runtime monitoring:
  • Where feasible, monitor for anomalous command executions from the web server process.
  • Use network intrusion detection systems to flag unusual HTTP requests to /cgi/iux_set.cgi containing shell metacharacters or suspicious patterns in the RestoreFile parameter.

These approaches help spot both existing issues and regressions in future firmware releases.

Sources

SourceLink
github.comgithub.com/glkfc/IoT-Vulnerability/blob/main/iptime/c200/sub_409054_vulnerability_report_EN.md
vuldb.comvuldb.com/submit/807786
vuldb.comvuldb.com/vuln/361112
vuldb.comvuldb.com/vuln/361112/cti

Frequently asked

questions

It can be, if the device's web management interface is exposed to the internet or reachable through port forwarding or cloud-access features. If access is restricted to a local network, an attacker would require a position on that network, but no physical access to the device is typically needed.

That depends on the specific deployment and how the ApplyRestore endpoint is protected. In some configurations, authentication may gate access to the vulnerable function, while in others misconfigurations or weak access control could allow unauthenticated requests. The essential requirement is the ability to send crafted HTTP requests to the vulnerable endpoint.

The most robust fix is to remove shell-based command construction entirely for restore operations and instead use direct APIs or libraries to manipulate configuration files and system state. Any remaining use of system commands should separate arguments from data (no string concatenation), enforce strict validation and allowlists, and ensure no untrusted input directly reaches a command interpreter.

Prevention relies on secure design and coding practices: avoid passing user input to shells, adopt parameterized interfaces for system operations, enforce centralized input validation and encoding, and institutionalize security reviews for web endpoints that perform sensitive actions. Combining these patterns with automated static analysis and CI-based security checks reduces the likelihood of new command injection flaws reaching production firmware.

Find out what's already
exploitable in your codebase.