#Error connecting to LangGraph API running in production mode

1 messages · Page 1 of 1 (latest)

hallow gazelle
#

Hi @tiny frost . When you say production mode, do you mean when the LangGraph is deployed to LangGraph Cloud?

And: are you using Copilot Cloud? Or self hosting the copilot runtime?

tiny frost
#

Both LanGraph and CopilotKit are self hosted. We get this error These are the available agents: [Research Agent (ID: fe096781-5601-53d2-b2f6-0d3403f7e9ca)] where as locally it connects

desert tree
#

Hey @tiny frost
Can you post your CopilotRuntime instantiation?
I assume you are using LangGraph with python FastAPI...?

tiny frost
#

Hi @desert tree ,
We are using LangGraph API as a self hosted container run with langgraph dev --port 8123 driven by langraph.json

        deploymentUrl: process.env.COPILOT_KIT_URL! || '',
        langsmithApiKey: process.env.LANGSMITH_API_KEY!,
        agents: [{ name: 'agent', description: 'Research agent' }],
    });

    const runtime = new CopilotRuntime({
        remoteEndpoints: [remoteEndpoint],
    });

    const { handleRequest } = copilotRuntimeNextJSAppRouterEndpoint({
        runtime,
        serviceAdapter: llmAdapter,
        endpoint: '/api/copilotkit',
    });```
tiny frost
#

We have found a solution to the issue