Tencent's WeKnora is an LLM-powered framework designed for deep document understanding and semantic retrieval. In versions prior to 0.2.5, when the Agent service is enabled, users can call a database query tool that suffers from insufficient backend validation of input queries. This leads to an SQL Injection vulnerability (CWE-89), where attackers can craft prompt-based inputs to bypass query restrictions and execute arbitrary SQL commands. Such injection can allow unauthorized reading of sensitive information, modification of data, or disruption of database availability. The vulnerability has a CVSS 3.1 score of 5.6, reflecting medium severity, with network attack vector, high attack complexity, no privileges or user interaction required, and impacts on confidentiality, integrity, and availability rated as low. The flaw was publicly disclosed on January 10, 2026, and has been patched in WeKnora version 0.2.5. No known exploits have been reported in the wild, but the nature of the vulnerability makes it a significant risk for affected deployments, especially where the Agent service is enabled and exposed to untrusted users.

The SQL Injection vulnerability can lead to unauthorized disclosure of sensitive data stored in the backend database, potentially including confidential documents or user information processed by WeKnora. Attackers might also alter or delete data, impacting data integrity, or cause denial of service by disrupting database operations. Since the vulnerability can be exploited remotely without authentication, any exposed instance of WeKnora with the Agent service enabled is at risk. This can undermine trust in the system, cause regulatory compliance issues, and lead to operational disruptions. Organizations relying on WeKnora for document understanding and semantic retrieval may face data breaches or service interruptions if the vulnerability is exploited.

1. Upgrade WeKnora to version 0.2.5 or later, where the vulnerability is patched. 2. Disable the Agent service if it is not required, reducing the attack surface. 3. Implement strict input validation and sanitization on all user-supplied inputs, especially those used in database queries. 4. Employ parameterized queries or prepared statements in the backend to prevent SQL Injection. 5. Restrict database permissions to the minimum necessary for the application to operate, limiting potential damage. 6. Monitor logs for unusual query patterns or repeated failed attempts that may indicate exploitation attempts. 7. Conduct regular security assessments and code reviews focusing on input handling and database interactions. 8. If feasible, deploy Web Application Firewalls (WAFs) with SQL Injection detection rules tailored to WeKnora's query patterns.