Yes—set the flow run’s state to AwaitingRetry via the Prefect client. That schedules a new retry attempt for that run. Async example:

from uuid import UUID
from datetime import datetime, timezone
from prefect.client.orchestration import get_client
from prefect.states import AwaitingRetry

flow_run_id = UUID("00000000-0000-0000-0000-000000000000")  # your flow run ID

async with get_client() as client:
    result = await client.set_flow_run_state(
        flow_run_id=flow_run_id,
        state=AwaitingRetry(
            scheduled_time=datetime.now(timezone.utc),  # or a future time
            message="Manual retry",
        ),
        force=True,  # bypasses orchestration guards if needed
    )

    if result.status.value == "ACCEPT":
        print("Retry scheduled.")
    else:
        print(f"Retry not accepted: {result.details.reason}")

Sync example:

from uuid import UUID
from datetime import datetime, timezone
from prefect.client.orchestration import get_client
from prefect.states import AwaitingRetry

flow_run_id = UUID("00000000-0000-0000-0000-000000000000")

with get_client(sync_client=True) as client:
    result = client.set_flow_run_state(
        flow_run_id=flow_run_id,
        state=AwaitingRetry(
            scheduled_time=datetime.now(timezone.utc),
            message="Manual retry",
        ),
        force=True,
    )

    if result.status.value == "ACCEPT":
        print("Retry scheduled.")
    else:
        print(f"Retry not accepted: {result.details.reason}")

Notes: - Use a timezone-aware

scheduled_time

(e.g.,

datetime.now(timezone.utc)

) for immediate retry, or set a future time to delay it. -

force=True

is often necessary to override orchestration rules that might otherwise reject the transition. The worker will pick up the run when it becomes due.