logo

Database

Lack of data validation In langchain

Description

LangChain vulnerable to code injection In LangChain through 0.0.131, the LLMMathChain chain allows prompt injection attacks that can execute arbitrary code via the Python exec() method.

Mitigation

Update Impact

Minimal update. May introduce new vulnerabilities or breaking changes.

Ecosystem
Component
Affected version
Patched versions