client_notebook

FutureHouse platform client usage example

import time

from futurehouse_client import FutureHouseClient, JobNames
from futurehouse_client.models import (
    RuntimeConfig,
    TaskRequest,
)
from ldp.agent import AgentConfig

Client instantiation

Here we use auth_type=AuthType.API_KEY to authenticate with the platform. Please log in to the platform and go to your user settings to get your API key.

client = FutureHouseClient(
    api_key="your-api-key",
)

Submit a task to an available futurehouse job

In the futurehouse platform, we refer to the deployed combination of agent and environment as a job. Submitting task to a futurehouse job is done by calling the create_task method, which receives a TaskRequest object.

task_data = TaskRequest(
    name=JobNames.from_string("crow"),
    query="What is the molecule known to have the smallest solubility in water?",
)
client.create_task(task_data)

while client.get_task().status != "success":
    time.sleep(5)
print(f"Task status: {client.get_task().status}")
print(f"Task answer: \n{client.get_task().formatted_answer}")

You can also pass a runtime_config to the create_task method, which will be used to configure the agent on runtime. Here, we will define a agent configuration and include it in the TaskRequest. This agent is used to decide the next action to take. We will also use the max_steps parameter to limit the number of steps the agent will take.

agent = AgentConfig(
    agent_type="SimpleAgent",
    agent_kwargs={
        "model": "gpt-4o",
        "temperature": 0.0,
    },
)
task_data = TaskRequest(
    name=JobNames.CROW,
    query="How many moons does earth have?",
    runtime_config=RuntimeConfig(agent=agent, max_steps=10),
)
client.create_task(task_data)

while client.get_task().status != "success":
    time.sleep(5)
print(f"Task status: {client.get_task().status}")
print(f"Task answer: \n{client.get_task().formatted_answer}")

Continue a job

The platform allows to ask follow-up questions to the previous job. To accomplish that, we can use the runtime_config to pass the task_id of the previous task.

Notice that create_task accepts both a TaskRequest object and a dictionary with keywords arguments.

task_data = TaskRequest(
    name=JobNames.CROW, query="How many species of birds are there?"
)

task_id = client.create_task(task_data)
while client.get_task().status != "success":
    time.sleep(5)
print(f"First task status: {client.get_task().status}")
print(f"First task answer: \n{client.get_task().formatted_answer}")

continued_task_data = {
    "name": JobNames.CROW,
    "query": (
        "From the previous answer, specifically, how many species of crows are there?"
    ),
    "runtime_config": {"continued_job_id": task_id},
}

continued_task_id = client.create_task(continued_task_data)
while client.get_task().status != "success":
    time.sleep(5)
print(f"Continued task status: {client.get_task().status}")
print(f"Continued task answer: \n{client.get_task().formatted_answer}")