Post

I am working on an exciting feature: local agents with skein-js. These will enable you to build and run your agents locally using Tauri or Electron, etc., eliminating the need for server management, all while utilising the same Agents Protocol. github.com/skein-js/ske... #typescript #langgraph

Running agents inside Tauri means the model calls leave from the user machine, so API keys either ship in the app or sit behind a tiny proxy. Which way is skein-js leaning?

Skein-js provides the primitives for exposing your LangGraph agents to the application, by implementing the agents protocol and providing a runtime, making it easier for you to provide agentic capabilities in your preferred frontend... Think threads, memories, persistence, etc.

0 reposts1 like

Accessing keys can be done in so many ways, and we don't impose, we support envs, but for other configs your app needs, but how you manage keys and access to the models - service account, oauth, secrets manager is upto you at LangGraph level, just below the skein-js layer

Leaving key handling to the LangGraph layer makes sense. Most teams already have a secrets manager or OAuth flow they trust, and a frontend runtime that picks one for them gets ripped out fast.