/

CVE-2026-73670

Severity:

HIGH RISK

(7.2)

SQL Injection (CWE-89) in CMS Admin Database Handler

CVE-2026-73670

Deferred

Vulnerability type

Category:

SQL Injection

CWE:

CWE-89

Impact:

Database Schema Disclosure and Sensitive Data Exposure

Severity:

HIGH RISK

(7.2)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

A SQL injection vulnerability in the CMS administrative database handler allows authenticated administrators to influence a SHOW COLUMNS FROM query through an unsanitized table_name parameter. Depending on database permissions and application behavior, an attacker may enumerate schema details, access system metadata, and extract sensitive credential data through secondary injection paths.

Affected context

The issue occurs in the CMS administrative database functionality, specifically around admin/db_data.php and its handling of the table_name GET or POST parameter. It becomes risky when authenticated administrative input is incorporated into a database statement without strict identifier validation or safe query construction, particularly where the database account can read schema metadata or sensitive application tables.

Exploit preconditions

  • Attack surface: Administrative database-management functionality
  • Required access: Valid authenticated administrator access, or another account capable of reaching the affected administrative endpoint
  • Trigger point: Processing of the table_name GET or POST parameter in the database-column inspection operation
  • Why exploitation is possible: Untrusted input is incorporated into a SHOW COLUMNS FROM statement without adequate validation or safe handling of dynamic database identifiers

Why this happens

The underlying bug is improper neutralization of special elements in SQL statements, matching CWE-89: Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection'). The application treats a request-controlled table identifier as trusted query structure instead of constraining it to an approved set of identifiers or safely separating data from SQL syntax. This can permit boolean-based, time-based, error-based, or metadata-oriented query manipulation.

Potential impact

  • An attacker may enumerate database schemas, tables, and columns
  • The vulnerability could expose database metadata from sources such as information_schema
  • Sensitive application records, including credential-related data, may be disclosed
  • Database availability could be affected through expensive or delayed queries
  • Compromised data may support account takeover or further administrative abuse

General mitigation guidance

Mitigation generally focuses on ensuring that dynamic database identifiers cannot be supplied directly from untrusted requests.

Common defensive patterns include:

  • Restricting table and column names to a server-side allowlist mapped to fixed identifiers
  • Using database APIs that safely represent identifiers where dynamic names are necessary
  • Avoiding string concatenation when constructing SQL statements
  • Applying strict input validation based on expected identifier formats and allowed values
  • Limiting the database account to the minimum read and metadata permissions required
  • Separating administrative database inspection from access to credential and other sensitive tables
  • Returning generic database errors and avoiding detailed query diagnostics in responses

How teams detect issues like this

Organizations commonly detect this class of issue through:

  • Static analysis rules for SQL construction involving request parameters
  • Code review checks for concatenated GET or POST values in database queries
  • Tests that verify dynamic identifiers are allowlisted and rejected when unexpected
  • Monitoring for unusual administrative database requests, repeated metadata enumeration, and abnormal query timing
  • Database audit policies that alert on unexpected access to schema catalogs or credential-related tables
  • Automated security review tools that identify unsanitized input reaching SQL execution APIs

Sources

SourceLink
github.comgithub.com/DevVaibhav07/VULN-POC/blob/main/Saurus.md
github.comgithub.com/sauruscms/Saurus-CMS-Community-Edition
vulncheck.comvulncheck.com/advisories/cms-admin-sql-injection-via-db-data-php-table-name-parameter
vulncheck.comvulncheck.com/blog/filerun-thumbnail-command-injection-rce

Frequently asked

questions

It may be exploitable through the CMS administrative interface when that interface is reachable and the attacker has the required authenticated privileges.

The described attack path requires authenticated administrative access, although the practical scope depends on the application's role permissions and access controls.

Treat table names as controlled identifiers rather than query data: map accepted names to fixed server-side values, use safe identifier APIs where appropriate, and enforce least-privilege database permissions.

Combine parameterized query practices with allowlisted dynamic identifiers, input validation, security-focused code review, database activity monitoring, and tests covering every path from request parameters to SQL execution.

Find out what's already
exploitable in your codebase.