LangChain

Snapback for LangChain

io.github.ra1labsworkx-wq/snapback

Diagnose why an AI agent failed and get the verified fix instantly. Free, no token.

client:LangChain transport:streamable-http tools:23

Install Snapback in LangChain

pip install langchain-mcp-adapters

from langchain_mcp_adapters.client import MultiServerMCPClient

client = MultiServerMCPClient({
        "snapback": {
            "transport": "streamable_http",
            "url": "https://api.snapback.sh/mcp",
        }
})

tools = await client.get_tools()

get_tools() returns the server's tools as LangChain tools, ready for an agent. LangChain docs

What LangChain can do once it is connected

Snapback in other clients

Snapback MCP server