Attackers are targeting the Python notebook Marimo

Due to ongoing attacks, software developers should update Marimo promptly.

listen Print view
A hand holding a digital red globe with many warning signs

(Image: tete_escape/Shutterstock.com)

1 min. read

Currently, attackers are targeting the Python notebook Marimo and exploiting a security vulnerability to execute malicious code.

Marimo is an integrated development environment for Python that combines code, results, visualizations, and documentation.

As indicated by a warning message, the authentication in the context of the WebSocket endpoint /terminal/ws is broken, and attackers can exploit the “critical” vulnerability (CVE-2026-39987) without logging in.

Subsequently, they gain access to a shell with far-reaching privileges and can execute their commands. Given the severity rating, it is assumed that systems will be considered fully compromised afterward. The extent to which the attacks are occurring and the targets of the hitherto unknown attackers are currently unclear.

Security researchers from Sysdig discovered these attacks. In a report, they state that twelve hours after the vulnerability became known, they observed attack activities from 125 IP addresses. They report that after successful access, attackers obtain credentials such as SSH keys and use them to spread further within networks. Therefore, for security reasons, Marimo users should change their database credentials and API keys to deny attackers access.

Videos by heise

The developers assure that the vulnerability has been closed in Marimo 0.23.0. The current release is 0.23.1. Those who cannot install the security update immediately must restrict access to /terminal/ws or temporarily disable the function.

(des)

Don't miss any news – follow us on Facebook, LinkedIn or Mastodon.

This article was originally published in German. It was translated with technical assistance and editorially reviewed before publication.