agents/outline_agent.py

from langchain_community.callbacks import get_openai_callback
from langchain_core.prompts import PromptTemplate
from langchain_openai import ChatOpenAI

from prompts.outline_agent import template


class OutlineAgent:
    def __init__(self, model: str = "gpt-4o", temperature: float = 0.3) -> None:
        self._prompt = PromptTemplate(input_variables=["query"], template=template)

        self._llm = ChatOpenAI(model=model, temperature=temperature)

        self._chain = self._prompt | self._llm

    def run(self, query: str) -> tuple[str, list[dict[str, str]], dict[str, int | float]]:
        with get_openai_callback() as cb:
            outline = self._chain.invoke({"query": query}).content.strip()

            tokens = cb.total_tokens
            cost = cb.total_cost

        elements = outline.split("\n\n")

        main_title = elements[0][elements[0].index(":") + 1 :].strip().replace('"', "")

        themes = []
        keys = ["title", "query", "description"]

        for theme in elements[1:]:
            items = theme.strip().split("\n")
            themes.append({keys[i]: item[item.index(":") + 1 :].strip().replace('"', "") for i, item in enumerate(items)})

        return main_title, themes, {"tokens": tokens, "cost": cost}