External risk intelligence

LMFDB Server-Side Template Injection Vulnerability.

CVE advisorySeverity: CRITICAL (CVSS 9.3)

CVE-2026-123198

The vulnerability exists in a web application endpoint (/knowledge/render/<ID>) that is designed to accept user-supplied input via request parameters or query strings. As this is a web-based application feature reachable by unauthenticated users, it is commonly deployed as an internet-facing service, making it likely to be exposed to public network traffic.

Halo Surface Signal: 4 out of 5 — likely to be public-facing.

External exposure likelihood

Horizon Alert

Summary of the vulnerability and why it matters

This advisory concerns a critical vulnerability in LMFDB, a web application technology, that could allow unauthorized command execution on the server. The issue stems from how the application processes certain data, potentially enabling attackers to compromise the system without needing any credentials.

Unauthenticated attackers can run commands on the server.
This could lead to unauthorized system access and control.
Confirm relevance and exposure to our LMFDB instances.

Attack Path

How an attacker could exploit the issue

An unauthenticated attacker can exploit this vulnerability by sending a specially crafted request to the `/knowledge/render/<ID>` endpoint. This request can contain malicious Jinja expressions within query-string values or the 'content' data, which are then processed by the `render_knowl` function. Successful injection allows the attacker to interact with Python builtins and potentially execute arbitrary operating system commands as the web application user.

Reaches endpoint via network.
Injects Jinja via request parameters.
Allows arbitrary command execution.

Live Threat

Current exploitation, exposure, and threat context

This vulnerability could impact the confidentiality, integrity, and availability of the web application and the underlying operating system. When supported by the advisory, unauthenticated attackers can inject malicious Jinja expressions through request parameters, allowing them to execute arbitrary operating system commands as the web application account. This could lead to unauthorized access, modification, or deletion of data, as well as disruption of the service.

Web application and OS commands at risk.
Injection via crafted request parameters or data.
System compromise and data loss.

Operational Fix

Recommended remediation, mitigation, and detection steps

Given the server-side template injection vulnerability in LMFDB, the application owners and platform teams are likely responsible for remediation. The first practical step is to identify all instances of LMFDB within the environment, confirm their exposure and business criticality, and then engage the accountable owners to plan and execute the necessary updates.

Application owners should manage remediation.
Verify external and business-critical instances first.
Plan maintenance and coordinate vendor support.

Supplementary metadata

CVSS vector

Validate whether this threat affects your internet-facing exposure.

Halo Threat Intelligence helps prioritize remediation with Halo Surface Signal and H/A/L/O context. Start exposure validation with a free external attack surface trial.

Free EASM Trial See How Signal Works

Frequently asked questions

What is LMFDB?

LMFDB is a web-based database and search engine used to catalog mathematical objects, such as L-functions, modular forms, and number fields. It provides a collaborative environment for researchers to share and explore complex mathematical data via a browser interface.

What does server-side template injection mean in CVE-2026-123198?

This vulnerability, classified as CWE-1336 (Improper Neutralization of Special Elements Used in a Template Engine), occurs when an application processes user input as part of a template rather than simple data. In this instance, the application incorrectly executes injected Jinja expressions, allowing unauthorized access to Python's internal functions and the underlying operating system.

How can an attacker trigger this vulnerability?

An attacker sends a specifically crafted web request to the /knowledge/render/<ID> endpoint. By placing malicious Jinja syntax into query-string parameters or the content field, they force the server to interpret the input as code. Standard requests that do not include these specific template expressions or target different application paths do not trigger this command execution.

Is my LMFDB instance at risk?

According to Halo Surface Signal, this vulnerability is highly relevant if your LMFDB instance is internet-facing. Because the affected /knowledge/render/<ID> endpoint is designed to process web requests and is reachable by unauthenticated users, any instance exposed to public network traffic is a potential target for remote attackers.

What should I do to secure my LMFDB environment?

First, locate all deployed instances of LMFDB in your infrastructure to determine which are internet-accessible. Prioritize these for review and work with the technical teams managing the software to plan an update. Verify you are using a version beyond commit 01829bd, which contains the fix for the render_knowl function.

References

Written by Nicholas Merritt

Nicholas Merritt is VP of Security Solutions at Halo Security. With more than 25 years across application security, vulnerability prioritization, and offensive security, he helps teams translate threat data into practical exposure validation and remediation focus. He authors Halo Threat Intelligence CVE advisories using Halo Surface Signal and H/A/L/O context to prioritize internet-facing risk.