CVE-2026-126199: pgEdge Postgres MCP Server write confirmation bypass in SQL classification

CVE-2026-126199 · Severity: medium · CVSS 6.8 · Published 2026-11-12

pgEdge Postgres MCP Server is a tool that allows AI assistants and applications to execute SQL queries against PostgreSQL databases. Prior to version 1.1.0, the server's write confirmation prompt could be bypassed by using certain SQL constructs (SELECT INTO, data-modifying CTEs, and EXPLAIN ANALYZE) that appeared to be read operations but actually modified data. An attacker with access to a client configured to allow writes could execute unconfirmed database modifications without user awareness.

Technologies: pgEdge Postgres MCP Server. Vendors: pgEdge.

Quick answers

Is CVE-2026-126199 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-126199 being exploited in the wild.
How severe is CVE-2026-126199?
Medium severity, with a CVSS score of 6.8, published 12 November 2026.
Which products and versions are affected?
pgEdge Postgres MCP Server before 1.1.0.
Is there a fix for CVE-2026-126199?
Yes: fixed in 1.1.0. Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.
How is CVE-2026-126199 attacked?
It is attacked over the network, in a software library.
Am I affected by CVE-2026-126199?
Check whether you run pgEdge Postgres MCP Server 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.

How to fix

  • Fixed in: 1.1.0

Technical details

The vulnerability stems from incomplete SQL statement classification in the CLI and web clients, which determined write permissions by examining only the first keyword of SQL statements. Malicious constructs like SELECT ... INTO, WITH clauses containing INSERT/UPDATE/DELETE, and EXPLAIN ANALYZE of write operations bypassed the confirmation prompt on allow_writes connections. The fix implements a more robust SQL parser that strips comments and literals before classification, detects write indicators regardless of first keyword, and requires ANALYZE on EXPLAIN to trigger write classification.

Affected products

  • pgEdge Postgres MCP Server before 1.1.0

Timeline

  • 2026-11-12: disclosed
  • 2026-08-07: patched

References

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