Langchain Agno Bridge - Enhanced Agent development

General
, , ,
1

Hello Agno community and the Agno team!

I am excited to share that I have created a way to give superpowers to Agno.

Basically have I enhanced Agno by combining Agno with Langchain. I created this bridge between these two open source projects by doing some changes to how Agno handles tools.

When Agno handles tools as functions it calls the function and you get an response, but what happens if you give this function a Langchain LLM and a Python class with Langchain implementations…

You get more thoughtful responses, as the level of control through in for example prompt engineering is more in depth.

Here is my tutorial/project:

Best Regards,
Marcus

2

Hey @marcusjihansson,
Thank you for sharing this with us. This looks great.

We will test it out.

Thank you again

3

Hello Monali,

Have you had time to test it out?

I have alos added more files to this github with avialable tools and Langchain functions.

But I came across a problem, I think, which is that there would also be possible to use DSPy to enhance the prompting inside this Agent class by adding a DSPy layer like this:

In this case has all the prompting been removed but instead been implemented by coding the backend class to do the RAG instead of using the Langchain type prompting.

Thank you for your time.

4

Hey @marcusjihansson,

Apologies for the delay—I haven’t had a chance to test this out yet. I’ve shared your request with the team, and we’re working through all inquiries one by one. We’ll get back to you as soon as possible.

Thank you for your patience and understanding!