Built a Rust package that connects to AI agents from any framework in just a few lines of code. Full support for streaming and non-streaming operations with both sync and async APIs. Built on Tokio with type-safe error handling, WebSocket streaming via futures::Stream, and zero-cost abstractions for production workloads.
I'm talking about the Letta part - unless I'm missing something, the RunAgent framework has to communicate with the Letta server over the Letta API (I'm saying that part can be done w/ Python SDKs, TS SDKs, REST APIs, and down the line Rust natively etc).
But question is how they are standardizing it? I understand the multi framework approach, but like if anyone is a javascript developer, he doesn't know about python. How he can write that agent file. That's a question where I am confused.
Guys this is really a very unique approach. Just one thing I am confused about, as far as I understand, this is like vercel but for ai agent with multi agentic framework. Is it correct?