Security brief

Historical advisory: LangChain prompt injection could trigger server-side requests

REVIEW

CVE-2023-32786 concerns attacker-influenced URL retrieval in older LangChain releases.

Published 29 Oct 2026 · Material update 11 Nov 2023 — last time the underlying evidence changed · Evidence: high

What happened

A reviewed October 2023 advisory describes server-side request forgery in LangChain before version 0.0.329. Prompt injection could cause the service to retrieve an attacker-chosen URL and potentially introduce content into downstream processing.

The relevant capability is network access from the service, rather than access from a user’s browser. Review which destinations the application can reach and whether untrusted input can influence retrieval.

The advisory identifies 0.0.329 as the historical fixed version. Update affected legacy dependencies to a supported patched release. As a separate defensive measure, restrict outbound access where the application does not need it. The source does not establish compromise of every application using LangChain.

Read the original source.

Who should check?

Operators of older LangChain applications performing URL retrieval.

Next step

Review the affected version and constrain unnecessary outbound access.

What we know

Historical fixed version
0.0.329

What remains unknown

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

Source references

Technical details

CVE
CVE-2023-32786