Build a sports research agent with LangChain and one REST tool

pip install langchain-core langchain-anthropic httpx
ANTHROPIC_MODEL=<your model id> python langchain_slate_tool.py

"""Slate research with LangChain: one REST tool, no framework magic. Set ANTHROPIC_API_KEY and ANTHROPIC_MODEL."""
import os

import httpx
from langchain_anthropic import ChatAnthropic
from langchain_core.messages import HumanMessage, SystemMessage, ToolMessage
from langchain_core.tools import tool

API = "https://deepchampai.com/api/agent/v1"
SYSTEM = (
    "You write short pre-game research briefings from DeepChamp tool results. Use only what the tools return. Say when the data was current (meta.as_of). Never state odds, picks or outcomes, and never claim a result. End with: Research only, not betting advice. 21+. Help: 1-800-MY-RESET."
)


@tool
def get_daily_slate(sport: str) -> str:
    """Today's games for a sport (nba, nfl, mlb, nhl, cfb, cbb or wnba) with start times and rest and schedule context. Returns JSON with data, compliance and meta."""
    response = httpx.get(f"{API}/slate/today", params={"sport": sport}, timeout=20)
    response.raise_for_status()
    return response.text


def run(question: str) -> str:
    llm = ChatAnthropic(model=os.environ["ANTHROPIC_MODEL"]).bind_tools([get_daily_slate])
    messages = [SystemMessage(SYSTEM), HumanMessage(question)]
    for _ in range(4):
        reply = llm.invoke(messages)
        messages.append(reply)
        if not reply.tool_calls:
            if isinstance(reply.content, str):
                return reply.content
            return "".join(block.get("text", "") for block in reply.content if isinstance(block, dict))
        for call in reply.tool_calls:
            messages.append(ToolMessage(content=get_daily_slate.invoke(call["args"]), tool_call_id=call["id"]))
    return "Stopped after four tool rounds."


if __name__ == "__main__":
    print(run("What should I check first on tonight's NBA slate?"))