Intel

AIKIDO-2026-783472

lightrag-hku is vulnerable to Server-Side Request Forgery (SSRF)

Server-Side Request Forgery (SSRF)GHSA-vv3m-f8x4-7377 Published 4 days ago

71

High Risk

This Affects:

PYTHONlightrag-hku
0.0.1 - 1.5.4
Fixed in 1.5.5
Are you affected? Scan for Free

TL;DR

LightRAG's native markdown parser in lightrag/parser/markdown/parser.py downloads external images referenced by uploaded markdown or textpack documents. The _validated_addresses SSRF guard resolves the image host and rejects non-global addresses, but evaluates the raw resolved address without decoding IPv6-transition wrappers such as NAT64 64:ff9b::/96, IPv4-compatible ::a.b.c.d, or 6to4 2002::/16 that embed an internal IPv4. On a host with NAT64/DNS64 routing, a wrapped address is classified as globally routable and the request is delivered to the embedded internal target, exposing loopback services, RFC1918 hosts, and cloud metadata endpoints. The fix decodes the embedded IPv4 and default-denies untrustworthy wrapper blocks before the global-routability check.

Who does this affect?

You are affected if you are using a version that falls within the vulnerable range and you allow document uploads processed by the native markdown parser with image download enabled on a host that has NAT64/DNS64 routing.

Background info

lightrag-hku is vulnerable to Server-Side Request Forgery (SSRF) in versions 0.0.1 - 1.5.4.

How to fix this

Upgrade the lightrag-hku library to the patch version.