Skip to main content
Background execution allows you to start an agent run that returns immediately with a PENDING status, while the actual work continues in the background. You can then poll for completion or cancel the run.
"""
Example demonstrating background execution with polling and cancellation.

Background execution allows you to start an agent run that returns immediately
with a PENDING status, while the actual work continues in the background.
You can then poll for completion or cancel the run.

Requirements:
- PostgreSQL running (./cookbook/scripts/run_pgvector.sh)
- OPENAI_API_KEY set

Usage:
    .venvs/demo/bin/python cookbook/02_agents/other/background_execution.py
"""

import asyncio

from agno.agent import Agent
from agno.db.postgres import PostgresDb
from agno.models.openai import OpenAIResponses
from agno.run.base import RunStatus

# ---------------------------------------------------------------------------
# Config
# ---------------------------------------------------------------------------

db = PostgresDb(
    db_url="postgresql+psycopg://ai:ai@localhost:5532/ai",
    session_table="background_exec_sessions",
)


# ---------------------------------------------------------------------------
# Create and Run Background Examples
# ---------------------------------------------------------------------------


async def example_background_run_with_polling():
    """Start a background run and poll until complete."""
    print("=" * 60)
    print("Example 1: Background run with polling")
    print("=" * 60)

    agent = Agent(
        name="BackgroundAgent",
        model=OpenAIResponses(id="gpt-5-mini"),
        description="An agent that runs in the background",
        db=db,
    )

    # Start a background run — returns immediately with PENDING status
    run_output = await agent.arun(
        "What is the capital of France? Answer in one sentence.",
        background=True,
    )

    print(f"Run ID: {run_output.run_id}")
    print(f"Session ID: {run_output.session_id}")
    print(f"Status: {run_output.status}")
    assert run_output.status == RunStatus.pending, (
        f"Expected PENDING, got {run_output.status}"
    )

    # Poll for completion
    print("\nPolling for completion...")
    for i in range(30):
        await asyncio.sleep(1)
        result = await agent.aget_run_output(
            run_id=run_output.run_id,
            session_id=run_output.session_id,
        )
        if result is None:
            print(f"  [{i + 1}s] Run not found in DB yet")
            continue

        print(f"  [{i + 1}s] Status: {result.status}")

        if result.status == RunStatus.completed:
            print(f"\nCompleted! Content: {result.content}")
            break
        elif result.status == RunStatus.error:
            print(f"\nFailed! Content: {result.content}")
            break
    else:
        print("\nTimed out waiting for completion")

    print()


async def example_cancel_background_run():
    """Start a background run and cancel it before completion."""
    print("=" * 60)
    print("Example 2: Cancel a background run")
    print("=" * 60)

    agent = Agent(
        name="CancellableAgent",
        model=OpenAIResponses(id="gpt-5-mini"),
        description="An agent whose run can be cancelled",
        db=db,
    )

    # Start a long background run
    run_output = await agent.arun(
        "Write a very detailed essay about the history of computing. "
        "Make it at least 5000 words with sections and subsections.",
        background=True,
    )

    print(f"Run ID: {run_output.run_id}")
    print(f"Status: {run_output.status}")

    # Wait a moment for the run to start
    await asyncio.sleep(2)

    # Cancel the run
    print("Cancelling run...")
    cancelled = await agent.acancel_run(run_id=run_output.run_id)
    print(f"Cancel result: {cancelled}")

    # Check the final state
    await asyncio.sleep(1)
    result = await agent.aget_run_output(
        run_id=run_output.run_id,
        session_id=run_output.session_id,
    )
    if result:
        print(f"Final status: {result.status}")
    print()


async def example_cancel_before_start():
    """Cancel a run before it even starts (cancel-before-start semantics)."""
    print("=" * 60)
    print("Example 3: Cancel-before-start")
    print("=" * 60)

    from agno.run.cancel import cancel_run

    agent = Agent(
        name="PreCancelAgent",
        model=OpenAIResponses(id="gpt-5-mini"),
        description="An agent whose run is cancelled before starting",
        db=db,
    )

    # Pre-generate a run ID
    from uuid import uuid4

    run_id = str(uuid4())

    # Cancel the run BEFORE it starts
    print(f"Pre-cancelling run {run_id}...")
    cancel_run(run_id)

    # Now start the run with that ID — it should detect the cancellation
    run_output = await agent.arun(
        "This should be cancelled before it runs.",
        background=True,
        run_id=run_id,
    )

    print(f"Run ID: {run_output.run_id}")
    print(f"Initial status: {run_output.status}")

    # Wait and check — the background task should detect the cancellation
    await asyncio.sleep(2)
    result = await agent.aget_run_output(
        run_id=run_output.run_id,
        session_id=run_output.session_id,
    )
    if result:
        print(f"Final status: {result.status}")
    print()


async def main():
    await example_background_run_with_polling()
    await example_cancel_background_run()
    await example_cancel_before_start()
    print("All examples completed!")


if __name__ == "__main__":
    asyncio.run(main())

Run the Example

# Clone and setup repo
git clone https://github.com/agno-agi/agno.git
cd agno/cookbook/02_agents/14_advanced

# Create and activate virtual environment
./scripts/demo_setup.sh
source .venvs/demo/bin/activate

# Optiona: Run PgVector (needs docker)
./cookbook/scripts/run_pgvector.sh

# Export relevant API keys
export OPENAI_API_KEY="***"

python background_execution.py