CVE-2026-79536: ByteBase DBHub read-only mode bypass via SQL comment in MySQL

CVE-2026-79536 · Severity: info · Published 2026-09-29

ByteBase DBHub is a database management tool that allows operators to safely expose databases to AI agents and clients in read-only mode. An attacker can bypass this read-only protection on MySQL and MariaDB backends by injecting hidden SQL statements using comment syntax, allowing them to write data, modify rows, or drop tables despite the read-only guarantee. No authentication is required for this exploit in the default configuration.

Technologies: Bytebase DBHub. Vendors: Bytebase.

Quick answers

Is CVE-2026-79536 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-79536 being exploited in the wild.
How severe is CVE-2026-79536?
Informational severity, published 29 September 2026.
Which products and versions are affected?
Bytebase DBHub v1.2.0 and earlier.
Is there a fix for CVE-2026-79536?
No fixed version is recorded yet. Check the vendor's advisory in the references below. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.
How is CVE-2026-79536 attacked?
It is attacked over the network, in a software library.
Am I affected by CVE-2026-79536?
Check whether you run Bytebase DBHub at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

Technical details

The vulnerability is an authorization bypass (CWE-863) and SQL injection (CWE-89) in the SQL classifier (src/utils/sql-parser.ts, function scanTokenMySQL) that fails to recognize # as a single-line comment in MySQL/MariaDB. An attacker can prefix a write or DDL statement with # to hide it from the classifier; the parser treats # as a string literal rather than a comment separator, so it sees only the leading harmless read statement and permits the request. When the unmodified SQL reaches MySQL/MariaDB with multipleStatements enabled, both statements execute, allowing writes or destructive DDL despite read-only mode. Network access to DBHub's HTTP endpoint is required, but no authentication is needed by default.

Affected products

  • ByteBase DBHub v1.2.0 and earlier

Timeline

  • 2026-07-14: disclosed: Vulnerability reported to maintainer by Traceforce
  • 2026-08-21: advisory: Reported via MITRE
  • 2026-09-10: other: CVE ID CVE-2026-79536 assigned by MITRE
  • 2026-09-22: disclosed: Security advisory published by Traceforce

References

This page as Markdown. Free to reuse under CC BY 4.0 with a link back.