/

CVE-2026-86189

Severity:

CRITICAL RISK

(9.8)

Path Traversal (CWE-73) in AVideo FFmpeg Notify Endpoint

CVE-2026-86189

Deferred

Vulnerability type

Category:

Path Traversal

CWE:

CWE-73

Impact:

Unauthenticated Arbitrary File Write

Severity:

CRITICAL RISK

(9.8)

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

WWBN AVideo contains a path traversal flaw in its FFmpeg notification endpoint, allowing unauthenticated requests to write files outside the intended directory. The risk arises when an attacker supplies a caller-controlled path and reuses an issued ciphertext as a notifyCode token that is decrypted but not properly validated.

Affected context

The vulnerability occurs in deployments exposing the notify.ffmpeg.json.php notification endpoint when the avideoRelativePath parameter influences filesystem paths without adequate confinement. It is especially concerning where the web application process can write to application directories or other sensitive locations, and where notifyCode values are accepted after decryption without authenticity or freshness validation.

Exploit preconditions

  • Attack surface: The publicly reachable FFmpeg notification endpoint
  • Required access: No authenticated account is required according to the vulnerability description
  • Trigger point: Processing of the avideoRelativePath parameter and notifyCode token
  • Why exploitation is possible: Caller-controlled path data is used for file placement without effective traversal prevention, while decrypted ciphertext is not validated as an authorized notification token

Why this happens

The underlying issue is external control of a file path without reliable canonicalization, directory confinement, or trusted-path enforcement. This maps to CWE-73: External Control of File Name or Path, because request-controlled data can influence where the application writes files. The authentication weakness compounds the problem: decryption alone is treated as sufficient token validation, allowing previously issued ciphertext to be replayed rather than requiring integrity, authorization, and freshness checks.

Potential impact

  • Unauthenticated attackers may write files to arbitrary locations permitted by the application process
  • Application files, configuration, or content may be modified or replaced
  • The web application may be defaced, disrupted, or made unavailable
  • Server-side code execution could occur if a writable location is interpreted as executable content
  • Persistent compromise or follow-on access may result

General mitigation guidance

Mitigation generally focuses on eliminating client control over filesystem destinations and ensuring that notification tokens provide authenticated authorization.

Common defensive patterns include:

  • Resolve paths against a fixed, trusted base directory and reject values that escape that directory
  • Prefer server-selected identifiers and filenames over request-supplied paths
  • Apply strict allowlists for permitted subdirectories and filename formats
  • Use authenticated, integrity-protected, scoped, and time-limited tokens rather than relying on decryption alone
  • Prevent token replay through nonce, expiration, or server-side state checks where appropriate
  • Run the application with the minimum filesystem permissions required
  • Keep upload and generated-content directories separate from executable application paths

How teams detect issues like this

Organizations can look for:

  • Static analysis findings where request parameters flow into filesystem APIs without canonicalization or base-directory checks
  • Code review signals such as string concatenation used to construct paths from HTTP parameters
  • Authentication logic that decrypts tokens but does not verify integrity, audience, scope, expiration, or replay status
  • Web and application logs showing unusual relative-path sequences, unexpected destination directories, repeated notification tokens, or unauthenticated write attempts
  • Policy enforcement that blocks application writes outside designated data directories
  • Automated security testing and code review tools that model path traversal and tainted data flows

Sources

SourceLink
github.comgithub.com/WWBN/AVideo/security/advisories/GHSA-cprx-fggj-7vpq
vulncheck.comvulncheck.com/advisories/wwbn-avideo-unauthenticated-path-traversal-via-notify-ffmpeg-json-php

Frequently asked

questions

Yes. The described endpoint accepts unauthenticated requests, so an attacker may be able to trigger the flaw remotely when the endpoint is reachable and the application process has relevant write permissions.

No normal user authentication is required according to the description. Network reachability to the notification endpoint and the ability to submit request parameters are the primary access conditions.

Do not derive filesystem destinations directly from client-controlled paths. Use server-side path selection, strict directory confinement, canonicalization, and authenticated, non-replayable notification tokens.

Apply least-privilege filesystem permissions, separate writable data from executable application directories, review all request-to-filesystem data flows, and enforce automated checks for path traversal and incomplete token validation.

Find out what's already
exploitable in your codebase.