Human in the Loop

This example shows how to implement human-in-the-loop functionality in your Agno tools. It shows how to:

• Add pre-hooks to tools for user confirmation
• Handle user input during tool execution
• Gracefully cancel operations based on user choice

Some practical applications:

• Confirming sensitive operations before execution
• Reviewing API calls before they’re made
• Validating data transformations
• Approving automated actions in critical systems

Code

human_in_the_loop.py

import json
from textwrap import dedent

import httpx
from agno.agent import Agent
from agno.tools import tool
from agno.utils import pprint
from rich.console import Console
from rich.prompt import Prompt

console = Console()


@tool(requires_confirmation=True)
def get_top_hackernews_stories(num_stories: int) -> str:
    """Fetch top stories from Hacker News.

    Args:
        num_stories (int): Number of stories to retrieve

    Returns:
        str: JSON string containing story details
    """
    # Fetch top story IDs
    response = httpx.get("https://hacker-news.firebaseio.com/v0/topstories.json")
    story_ids = response.json()

    # Yield story details
    all_stories = []
    for story_id in story_ids[:num_stories]:
        story_response = httpx.get(
            f"https://hacker-news.firebaseio.com/v0/item/{story_id}.json"
        )
        story = story_response.json()
        if "text" in story:
            story.pop("text", None)
        all_stories.append(story)
    return json.dumps(all_stories)


# Initialize the agent with a tech-savvy personality and clear instructions
agent = Agent(
    description="A Tech News Assistant that fetches and summarizes Hacker News stories",
    instructions=dedent("""\
        You are an enthusiastic Tech Reporter

        Your responsibilities:
        - Present Hacker News stories in an engaging and informative way
        - Provide clear summaries of the information you gather

        Style guide:
        - Use emoji to make your responses more engaging
        - Keep your summaries concise but informative
        - End with a friendly tech-themed sign-off\
    """),
    tools=[get_top_hackernews_stories],
    markdown=True,
)

# Example questions to try:
# - "What are the top 3 HN stories right now?"
# - "Show me the most recent story from Hacker News"
# - "Get the top 5 stories (you can try accepting and declining the confirmation)"
response = agent.run("What are the top 2 hackernews stories?")
if response.is_paused:
    for tool in response.tools:  # type: ignore
        # Ask for confirmation
        console.print(
            f"Tool name [bold blue]{tool.tool_name}({tool.tool_args})[/] requires confirmation."
        )
        message = (
            Prompt.ask("Do you want to continue?", choices=["y", "n"], default="y")
            .strip()
            .lower()
        )

        if message == "n":
            break
        else:
            # We update the tools in place
            tool.confirmed = True

    run_response = agent.continue_run(run_response=response)
    pprint.pprint_run_response(run_response)

Usage

1. Create a virtual environment

   Open the Terminal and create a python virtual environment.

   Mac

   python3 -m venv .venv
   source .venv/bin/activate
   
2. Install libraries

   pip install openai agno
   
3. Run the agent

   python human_in_the_loop.py