MCP Prompts and Resources

OSS Product Help LangGraph
, ,
1

Has anyone used MCP Prompts and Resources with LangGraph Agents?

I’m interested in learning about practical experiences or best practices for integrating MCP (Model Context Protocol) prompts and resources with LangGraph agents. I’ve seen examples of connecting MCP servers as tool providers for LangGraph agents, but I’m specifically curious about leveraging MCP’s prompt and resource features—not just tool execution.

  • Have you successfully used MCP prompts/resources in this setup?
  • Are there any limitations or tips to be aware of?
  • Any code snippets or references would be appreciated!

Thanks in advance for any insights.

Marcelo

2

You can integrate with prompts and resources (as well as tools) using the langchain-mcp-adapters package.

Will leave it to the community to share any usage patterns!

3

I, too, am very interested in this topic. I am currently trying to use the GitHub MCP to read the contents of a file using LangChain. Although the tool call works and I can see it returning the contents as an embedded resource, I am unable to figure out how to actually consume it and add it to my context. I’m wondering if anyone else has had any success with this.

My code is very concise to demonstrate the issue I’m running into, but I don’t want to spam it here unless others find it interesting. Let me know!