Security brief

Historical advisory: LangChain prompt loading could execute untrusted code

REVIEW

A 2023 advisory concerns LangChain prompt loading. Review legacy deployments rather than treating it as a new incident.

Published 29 Oct 2026 · Material update 7 Dec 2024 — last time the underlying evidence changed · Evidence: high

What happened

This historical advisory, first published in June 2023, concerns arbitrary code execution through LangChain’s load_prompt functionality. OSV identifies CVE-2023-34541 and lists 0.0.247 as the fixed package version.

The practical concern is the trust boundary around a configuration loader. Treating a prompt or configuration as passive data is unsafe when its loading path can execute code. That lesson remains useful even though the advisory is old.

For an inherited application, check the installed dependency and how prompt configurations enter the system. A minimum fix for this advisory is not a recommendation to deploy an obsolete release today. Choose a supported version after checking compatibility and other advisories.

Read the original source.

Who should check?

Maintainers of legacy LangChain applications using prompt loading.

Next step

Compare installed packages with the advisory and migrate to a supported patched release.

What we know

First disclosed
June 2023
Advisory fixed version
0.0.247

What remains unknown

Open question
The sources reviewed do not establish whether your deployment was exploited.

Source references

Technical details

CVE
CVE-2023-34541