What is the CVSS score of CVE-2026-32628?

CVE-2026-32628 has a CVSS 4.0 base score of 7.7 (High). CVSS vector: CVSS:4.0/AV:N/AC:H/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N. EPSS exploitation probability: 0.0%.

Which versions of PostgreSQL are affected by CVE-2026-32628?

AnythingLLM versions 1.11.1 and earlier contain a critical SQL injection vulnerability in the SQL Agent plugin that allows any user with agent access to execute arbitrary commands against connected…

How to fix or mitigate CVE-2026-32628?

Within 24 hours: Disable the SQL Agent plugin in all AnythingLLM instances and audit logs for unauthorized SQL execution. Within 7 days: Identify all connected databases and implement network-level access restrictions; conduct forensic review of database activity logs for compromise indicators. Within 30 days: Upgrade to a patched version when available; implement database user accounts with least-privilege SQL permissions; deploy Web Application Firewall (WAF) rules to detect SQL injection patterns if SQL Agent must remain enabled.

PostgreSQL CVE-2026-32628

EUVD-2026-12138 HIGH

SQL Injection (CWE-89)

2026-03-13 GitHub_M

SQLi PostgreSQL MySQL MSSQL Information Disclosure AI / ML Anything Llm

7.7

CVSS 4.0 · GitHub Advisory

Severity by source

GitHub Advisory PRIMARY

7.7 HIGH

CVSS:4.0/AV:N/AC:H/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

Primary rating from GitHub Advisory · only source for this CVE.

CVSS VectorGitHub Advisory

CVSS:4.0/AV:N/AC:H/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

Attack Vector

Network

Attack Complexity

High

Privileges Required

Low

User Interaction

None

Lifecycle Timeline

EUVD ID Assigned

Mar 13, 2026 - 22:01 euvd

EUVD-2026-12138

Analysis Generated

Mar 13, 2026 - 22:01 vuln.today

CVE Published

Mar 13, 2026 - 20:50 nvd

HIGH 7.7

DescriptionGitHub Advisory

AnythingLLM is an application that turns pieces of content into context that any LLM can use as references during chatting. In 1.11.1 and earlier, a SQL injection vulnerability in the built-in SQL Agent plugin allows any user who can invoke the agent to execute arbitrary SQL commands on connected databases. The getTableSchemaSql() method in all three database connectors (MySQL, PostgreSQL, MSSQL) constructs SQL queries using direct string concatenation of the table_name parameter without sanitization or parameterization.

AnalysisAI

SQL injection in AnythingLLM versions 1.11.1 and earlier enables authenticated users to execute arbitrary SQL commands against connected PostgreSQL, MySQL, and MSSQL databases through the built-in SQL Agent plugin. The vulnerability stems from unsafe string concatenation of table names in the getTableSchemaSql() method across all three database connectors, bypassing proper parameterization. Any user with access to invoke the SQL Agent can exploit this to read, modify, or delete sensitive database contents.

Technical ContextAI

AnythingLLM is a Large Language Model (LLM) context management application that allows users to convert content into contextual references for AI chatbots. The vulnerability exists in the getTableSchemaSql() method across all three database connector implementations (MySQL, PostgreSQL, and MSSQL), where table names are concatenated directly into SQL queries without proper parameterization or sanitization. This is a classic SQL injection vulnerability (CWE-89), where untrusted input flows directly into dynamically constructed SQL statements, allowing attackers to break out of the intended query structure and execute arbitrary SQL commands.

RemediationAI

Upgrade AnythingLLM to a version newer than 1.11.1 as soon as possible to address this SQL injection vulnerability. Until patching is complete, consider temporarily disabling the SQL Agent plugin functionality or restricting access to only highly trusted users who require this capability. Implement database activity monitoring to detect potential SQL injection attempts, and ensure database accounts used by AnythingLLM follow the principle of least privilege. Review and audit any recent SQL Agent usage logs for suspicious queries that may indicate exploitation attempts.