/

CVE-2026-93371

Severity:

HIGH RISK

(8.3)

Command Injection (CWE-77) in yt-dlp-web-ui Generic Download

CVE-2026-93371

Deferred

Vulnerability type

Category:

Command Injection

CWE:

CWE-74, CWE-77

Impact:

Remote Code Execution

Severity:

HIGH RISK

(8.3)

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 command injection vulnerability in yt-dlp-web-ui allows attacker-controlled download parameters to influence commands executed by the generic downloader.

Because the vulnerable functionality can be reached remotely in exposed deployments, an attacker may execute operating-system commands with the privileges of the application process.

Affected context

The vulnerability occurs in the NewGenericDownload function within server/internal/downloaders/generic.go of yt-dlp-web-ui deployments up to version 4. It becomes risky when externally supplied params values reach command construction or process execution without sufficient separation between data and executable arguments.

Exploit preconditions

  • Attack surface: Remote download functionality exposed through the yt-dlp-web-ui application
  • Required access: Ability to submit or influence download parameters; authentication requirements depend on deployment configuration
  • Trigger point: Processing a request by NewGenericDownload
  • Why exploitation is possible: User-controlled parameter data may be incorporated into an operating-system command without reliable command and argument boundary enforcement

Why this happens

The underlying bug is improper neutralization of special elements in externally controlled input used to construct an operating-system command. This maps primarily to CWE-77: Command Injection, with CWE-74: Improper Neutralization of Special Elements in Output Used by a Downstream Component describing the broader failure to treat command-interpreted data as untrusted.

Potential impact

  • An attacker may execute arbitrary commands on the application host
  • The attacker could access, modify, or delete files available to the service account
  • Sensitive application data or credentials may be exposed
  • The service could be disrupted or used for unauthorized outbound activity
  • A compromised process could enable lateral movement when excessive permissions or network access are available

General mitigation guidance

Mitigation generally focuses on ensuring that user-controlled values are never interpreted as shell syntax.

Common defensive patterns include:

  • Prefer direct process-execution APIs that pass an executable and discrete argument list without shell interpretation
  • Apply strict allowlists and type validation to download parameters
  • Separate command construction from request parsing and reject unexpected fields or formats
  • Avoid shell wrappers, concatenated command strings, and implicit evaluation of user input
  • Run the downloader with a dedicated least-privilege account and restrict filesystem and network permissions
  • Apply sandboxing or container isolation where practical, while treating isolation as a secondary control

How teams detect issues like this

Detection can combine static analysis, code review, runtime monitoring, and policy enforcement.

  • Search for shell execution APIs, command-string concatenation, interpolation, and subprocess calls that receive request-derived values
  • Review whether process APIs receive a discrete argument array or a single shell-parsed string
  • Trace HTTP or API parameters through validation, command construction, and execution boundaries
  • Add CI policies that flag shell invocation and untrusted data flows into process execution
  • Monitor for unexpected child processes, unusual command-line arguments, and outbound connections from the downloader service
  • Automated code review and data-flow analysis tools may help identify command-injection paths before deployment

Sources

SourceLink
github.comgithub.com/marcopiovanello/yt-dlp-web-ui/
github.comgithub.com/marcopiovanello/yt-dlp-web-ui/commit/c7ad3bd79c7c520a7d17e7f2ba19d962be8e7897
vuldb.comvuldb.com/cve/CVE-2026-93371
vuldb.comvuldb.com/submit/942846
vuldb.comvuldb.com/vuln/406851
vuldb.comvuldb.com/vuln/406851/cti

Frequently asked

questions

Yes, deployments that expose the affected download functionality may be remotely exploitable when an attacker can control the relevant parameters.

Not necessarily. The required access depends on the application's authentication, authorization, and network exposure settings.

Use direct process-execution interfaces with fixed executables and separately validated arguments, avoiding shell interpretation entirely.

Enforce strict input allowlists, review all data flows into process execution, apply least privilege, and use CI policies and runtime monitoring for command-injection patterns.

Find out what's already
exploitable in your codebase.