CVE-2026-114108 in Vibe-Trading
Summary
by MITRE • 10/22/2026
HKUDS Vibe-Trading from 0.1.0 before 0.1.7 contains a missing authentication vulnerability that allows unauthenticated attackers to access the FastAPI surface because require_auth() returns early when API_AUTH_KEY is unset. Attackers can post messages to /sessions/{id}/messages to trigger root shell commands via BashTool, upload .py or .sh files via /upload, read session histories, and obtain partial API keys.
If you want to get the best quality for vulnerability data then you always have to consider VulDB.
Analysis
by VulDB Data Team • 10/22/2026
The vulnerability identified in HKUDS Vibe-Trading versions prior to 0.1.7 represents a critical failure in access control mechanisms within the application's authentication layer. The core technical flaw stems from the implementation of the require_auth function, which is designed to enforce security checks before allowing access to protected endpoints. However, when the environment variable API_AUTH_KEY remains unset or empty during runtime initialization, this function fails to perform its intended validation and instead returns early without raising an exception or redirecting the request. This logic error effectively disables authentication for all routes that depend on this middleware, creating a state where the application operates in an unauthenticated mode by default if not explicitly configured with a secret key. Consequently, any user interacting with the FastAPI surface can bypass security controls entirely, treating protected endpoints as public resources.
This lack of authentication allows malicious actors to interact directly with sensitive internal APIs without providing valid credentials. Specifically, attackers can exploit this flaw to post messages to session-specific message endpoints such as /sessions/{id}/messages. Because these sessions are managed by the application and often involve automated agents or tools, the input provided through these endpoints is processed by backend logic that may invoke system-level commands. The vulnerability enables unauthenticated users to inject payloads into BashTool operations, which execute shell commands with elevated privileges on the host server. This capability transforms a simple authentication bypass into a severe Remote Code Execution (RCE) risk, as attackers can dictate arbitrary command-line arguments that are executed by the application process running under potentially privileged system accounts.
Beyond remote code execution, the absence of access control exposes significant data leakage risks and further exploitation vectors. Attackers can read session histories to gather contextual information about previous interactions, which may contain sensitive business logic details or internal state representations. Additionally, the vulnerability permits file uploads through endpoints like /upload, allowing attackers to place malicious Python scripts (.py) or shell scripts (.sh) onto the server filesystem. These files could be used for persistence, lateral movement within a networked environment, or as additional vectors for code execution if other application features allow their inclusion or sourcing. The ability to obtain partial API keys further compounds the risk by potentially enabling attackers to impersonate legitimate services or escalate privileges in connected systems that rely on those tokens for inter-service communication.
From an industry standard perspective, this vulnerability aligns with CWE-287 Improper Authentication and CWE-306 Missing Authentication for Critical Function. The ability to execute arbitrary commands via user-controlled input is a classic instance of CWE-78 OS Command Injection, while the unauthorized access to session data reflects CWE-200 Exposure of Sensitive Information to an Unauthorized Actor. In terms of attack tactics, this scenario maps directly to MITRE ATT&CK techniques including T1190 Exploit Public-Facing Application and potentially T1053 Scheduled Task/Job if uploaded scripts are configured for periodic execution. The early return in the authentication middleware is a common pattern error that underscores the importance of fail-secure design principles, where any ambiguity or missing configuration should result in denial of access rather than granting it.
Mitigation strategies must prioritize immediate remediation and robust security hygiene. The primary fix involves updating HKUDS Vibe-Trading to version 0.1.7 or later, which addresses the logic error in the require_auth function by ensuring that authentication is enforced regardless of environment variable states unless explicitly disabled for development purposes with clear warnings. For systems unable to update immediately, administrators should ensure that API_AUTH_KEY is always set to a strong, random value before deployment. Additionally, implementing defense-in-depth measures such as input validation on all user-supplied data passed to shell commands can mitigate the risk of command injection even if authentication bypasses occur. Network-level controls like web application firewalls (WAF) should be configured to detect and block patterns associated with OS command injection payloads in API requests. Regular security audits and static code analysis tools that flag missing error handling or improper access control checks are essential for preventing similar vulnerabilities in future development cycles.