GuardRails-AI
Guardrails AI is a framework to secure Large Language Model (LLM) applications from various threats such as jailbreaks and prompt injection etc. it is developed by Shreya Rajpal and her co-founders, it is mainly focusing on providing enterprise security for LLM applications using Reliable AI Markup Language (RAILS) which is language agnostic and the features and the configurations be ported over using these files, it is also programmable in the language itself, for now which is python , they have many pre-built guards available on Guardrails Hub.
It also has support for the LangChain framework to some extent through the Guard Rail Output Parser
Guardrails-AI looks promising
Guardrails AI is just a python package that can be integrated with any python application which removes the necessity to be confined to a specific framework or type of workflow such as NeMo Guardrails - NVIDIA
Although this is debatable as I have yet to work with them closely to come to a concrete conclusion