CVE-2026-121450 in CFNext
Summary
by MITRE • 11/03/2026
CFNext before 3.1.0 contains a weak cryptographic derivation vulnerability that allows subscription link holders to recover the admin password because the subscription UUID is an MD5 of the password with a fixed public salt. Attackers can brute-force the password offline without rate limiting, log into the admin panel, and read stored Trojan, Clash, Telegram and Cloudflare API secrets.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 11/03/2026
The vulnerability identified in CFNext versions prior to 3.1.0 represents a critical failure in cryptographic key derivation practices that fundamentally compromises the integrity of administrative authentication mechanisms. The core technical flaw lies in the method used to generate subscription Universally Unique Identifiers (UUIDs). Instead of employing a secure, salted hash function with appropriate computational cost parameters such as bcrypt or Argon2, the system utilizes MD5, an algorithm widely recognized for its cryptographic weaknesses and susceptibility to collision attacks. More critically, this implementation pairs the weak hashing algorithm with a fixed public salt derived from static configuration values rather than unique per-user random data. This design choice eliminates any entropy that might otherwise protect against pre-computed rainbow table attacks or brute-force enumeration, effectively reducing the security of password storage to simple reversible encryption given sufficient computational resources and time.
From an operational perspective, this architectural weakness allows attackers who possess valid subscription links to perform offline password recovery with negligible effort. Because the salt is public and fixed, any attacker can construct a lookup mechanism that maps MD5 hashes directly back to their plaintext equivalents without needing network interaction or triggering rate-limiting protections on the target server. This capability transforms what should be an authenticated access control issue into a straightforward data retrieval exercise for adversaries who have obtained subscription identifiers through legitimate means such as phishing, account sharing, or public exposure of API endpoints. The absence of rate limiting further exacerbates this risk by allowing unlimited attempts to test password guesses against the known hash structure until a match is found.
The impact of exploiting this vulnerability extends far beyond simple administrative access. Once an attacker recovers the admin credentials and logs into the control panel, they gain unrestricted visibility into sensitive operational secrets stored within the system. These include Trojan proxy configurations, Clash client settings, Telegram bot tokens, and Cloudflare API keys. The exposure of these specific artifacts poses severe risks to both service providers and their end-users. Compromised Trojan and Clash configurations can be used by attackers to set up unauthorized proxy nodes for malicious traffic routing, censorship circumvention, or further network infiltration. Exposed Telegram secrets allow for the takeover of bot accounts which may be used for spamming, phishing campaigns, or command-and-control operations. Similarly, leaked Cloudflare API keys provide potential access to DNS management and web application firewall configurations, enabling attackers to redirect traffic, intercept communications, or disrupt service availability entirely.
This vulnerability aligns with several established industry standards regarding security flaws. It is classified under CWE-798 as the use of hardcoded credentials, specifically due to the fixed salt mechanism that effectively hardcodes a known transformation key for all users. Additionally, it falls under CWE-328 which describes the use of weak cryptographic algorithms like MD5 where stronger alternatives are available and necessary. In terms of offensive security tactics, this flaw facilitates Initial Access through valid accounts obtained via credential stuffing or brute force techniques documented in MITRE ATT&CK technique T1078. The subsequent access to stored secrets constitutes Credential Access as defined by MITRE ATT&CK sub-technique T1552, where attackers retrieve unsecured credentials from local systems or configuration files.
Mitigation strategies must focus on immediate remediation of the cryptographic implementation and broader security hardening measures. Organizations running CFNext versions before 3.1.0 should upgrade to the latest stable release immediately as this version contains patches addressing these specific weaknesses. For environments where upgrading is not immediately feasible, administrators should rotate all exposed secrets including API keys, bot tokens, and proxy configurations assuming they are compromised due to the public nature of the salt used in password hashing. Implementing multi-factor authentication for administrative panels can provide a secondary layer of defense even if passwords are recovered through offline brute-forcing. Furthermore, future development efforts must prioritize using modern key derivation functions such as Argon2id or bcrypt with unique per-user salts and appropriate work factors to ensure that credential storage remains resilient against both online and offline attacks. Regular security audits focusing on authentication flows and secret management practices will help prevent similar vulnerabilities from being introduced in subsequent updates.