Async Tool Confirmation Required

This example demonstrates how to implement human-in-the-loop functionality with async agents, requiring user confirmation before executing tool operations.

Code

confirmation_required_async.py

"""🤝 Human-in-the-Loop: Adding User Confirmation to Tool Calls

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

Run `pip install openai httpx rich agno` to install dependencies.
"""

import asyncio
import json

import httpx
from agno.agent import Agent
from agno.models.openai import OpenAIChat
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)


agent = Agent(
    model=OpenAIChat(id="gpt-5-mini"),
    tools=[get_top_hackernews_stories],
    markdown=True,
)

run_response = asyncio.run(agent.arun("Fetch the top 2 hackernews stories"))
if run_response.is_paused:
    for tool in run_response.tools_requiring_confirmation:
        # 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":
            tool.confirmed = False
        else:
            # We update the tools in place
            tool.confirmed = True


run_response = asyncio.run(agent.acontinue_run(run_response=run_response))
# Or
# run_response = asyncio.run(agent.acontinue_run(run_id=run_response.run_id))

pprint.pprint_run_response(run_response)


# Or for simple debug flow
# asyncio.run(agent.aprint_response("Fetch the top 2 hackernews stories"))

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 -U agno openai httpx rich
   
3. Export your OpenAI API key

   Mac/Linux

     export OPENAI_API_KEY="your_openai_api_key_here"
   
4. Create a Python file

   Create a Python file and add the above code.

   touch confirmation_required_async.py
   
5. Run Agent

   Mac

   python confirmation_required_async.py
   
6. Find All Cookbooks

   Explore all the available cookbooks in the Agno repository. Click the link below to view the code on GitHub:

   Agno Cookbooks on GitHub