A critical SQL injection vulnerability in Tencent WeKnora allows unauthenticated attackers to bypass query restrictions and exfiltrate sensitive data from the backend database.

This vulnerability is an improper neutralization of special elements used in an SQL command (CWE-89). When the Agent service is enabled, the application lacks sufficient backend validation, allowing an unauthenticated attacker to inject malicious prompts to manipulate database queries.

The ability to perform unauthorized SQL queries poses a severe risk to data confidentiality and integrity. Attackers can extract sensitive information stored in the connected database, potentially leading to a significant data breach or unauthorized system control. While the CVSS score of 8.1 reflects a high risk, the exposure of backend database tools in an LLM-powered framework necessitates immediate attention to prevent intellectual property loss.

Immediate Action: Update Tencent WeKnora to version 0.2.5 or later to implement the necessary backend validation and neutralize the SQL injection vector.

Proactive Monitoring: Review application and database logs for unusual query patterns, such as unexpected SQL syntax or high volumes of data retrieval requests originating from the Agent service.

Compensating Controls: Deploy a Web Application Firewall (WAF) with updated rulesets to detect and block common SQL injection patterns and suspicious prompt-based payloads directed at the Agent service endpoint.

Public Exploit Available: Unknown

Given the potential for unauthorized data access and the existence of proof-of-concept evidence, organizations utilizing the WeKnora framework must prioritize this update. Administrators should verify their current deployment version immediately and apply the 0.2.5 patch to eliminate the SQL injection vulnerability, ensuring that the Agent service is properly secured against prompt-based bypass attempts.