arugulanet

Find offers that fit a request through Arugula, for Python agents. It uses an anonymous session, so no account is needed. Every offer comes with a tracked link the user follows to buy. The core has no dependencies.

Arugula is a matching engine: offers are ranked by fit only. The tools return a disclosure with the offers; show it to the user with them: "Ranked by fit only. Arugula may earn a commission if you buy through these links, at no extra cost to you."

By using Arugula you agree to arugulanet.com/terms. Privacy: arugulanet.com/privacy.

from arugulanet import Arugula

offers = Arugula().find_offers("a hotel room in Singapore for two nights", party_size=2)

LangChain and LangGraph

Install with pip install "arugulanet[langchain]", or "arugulanet[langgraph]". The tool sends channel langchain.

from arugulanet.langchain import find_offers_tool

tools = [find_offers_tool()]  # also for LangGraph: create_react_agent(model, tools)

CrewAI

Install with pip install "arugulanet[crewai]". The tool sends channel crewai.

from arugulanet.crewai import FindOffersTool

agent = Agent(role="Shopper", goal="Find what the user needs", backstory="...", tools=[FindOffersTool()])

More

• Example: python examples/agent.py "your request". Set ARUGULA_API_URL to use another API.
• Build: python -m build.
• Publish: twine upload dist/*, with the PyPI organization account.