Great question — this is a perfect use-case for an Automation in Prefect 3.x. High-level approach - Create two deployments of the same flow: - Deployment B: scheduled, uses parameters_b - Deployment A: no schedule, uses parameters_a (or accept overrides) - Create an Automation that listens for Completed events from Deployment B and runs Deployment A. Why two deployments? Event triggers cannot filter on specific parameter values. If you trigger the same deployment again, you’ll create a loop (your params_a run would also emit a Completed event that retriggers the automation). Using a second deployment avoids that. UI steps 1. Ensure you have two deployments of the same flow: - “my-flow - mode-b” (scheduled, default params_b) - “my-flow - mode-a” (no schedule, default params_a) 2. Go to Automations → New Automation 3. Trigger: - Type: Event - Posture: Reactive - Expected events:

prefect.flow-run.Completed

- Match related resource → Role: Deployment → pick “my-flow - mode-b” 4. Action: - Type: Run a deployment - Source: Selected - Deployment: “my-flow - mode-a” - Parameters (optional): if you didn’t set defaults on deployment A, add params_a here 5. Save and enable Python example

from datetime import timedelta
from uuid import UUID
from prefect.automations import Automation
from prefect.events.schemas.automations import EventTrigger, Posture, ResourceSpecification
from prefect.events.actions import RunDeployment

# IDs of the two deployments (same flow, different configs)
deployment_b_id = UUID("aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa")  # scheduled, params_b
deployment_a_id = UUID("bbbbbbbb-bbbb-bbbb-bbbb-bbbbbbbbbbbb")  # triggered, params_a

Automation(
    name="Run params_a after params_b completes",
    enabled=True,
    trigger=EventTrigger(
        expect={"prefect.flow-run.Completed"},
        posture=Posture.Reactive,
        match_related=ResourceSpecification({
            "prefect.resource.id": f"prefect.deployment.{deployment_b_id}"
        }),
        threshold=1,
        within=timedelta(seconds=0),
    ),
    actions=[
        RunDeployment(
            source="selected",
            deployment_id=deployment_a_id,
            # Optional if deployment A already has defaults:
            parameters={
                # put your params_a here
                # "foo": "bar"
            },
        )
    ],
).create()

Notes and tips - Filtering on parameter values isn’t supported in Event triggers; matching the specific deployment is the reliable filter. - If you want to pass values from the upstream flow run into the downstream run, you can use Jinja templates in the action’s parameters (e.g.,

{{ flow_run.parameters['x'] }}

), but this doesn’t help with filtering — only with parameterizing the action. Docs - Automations overview: Automations - Chaining deployments via events: Chaining deployments with events - Using parameters in action templates: Access parameters in templates Want me to draft the automation for you if you share the two deployment IDs (or names) and the params_a you want to send?

Short answer: not directly. In Prefect 3.x, Event-triggered automations can’t filter flow-run events by the originating schedule’s slug/name. Flow-run events do not expose the schedule as a related resource that you can match on, and EventTriggers can’t inspect payload fields like

created_by

to apply a condition. What you can filter on: - Deployment (recommended), flow, work pool/queue, tags, and flow-run name/id via resource labels. - You cannot filter by schedule slug or parameters. Practical workarounds - Easiest: create a separate deployment for that schedule. Then your automation can match on that deployment and react only to runs from that schedule. - If you must keep multiple schedules on one deployment: use a webhook action to do conditional logic server-side. - Automation: trigger on Completed for the deployment and CallWebhook to your service. - Your service: fetch the flow run by ID, check

flow_run.created_by.type == "SCHEDULE"

and

created_by.display_value

(which is the schedule slug if set). Only then trigger the follow-up run. Example automation using a webhook

from datetime import timedelta
from prefect.automations import Automation
from prefect.events.schemas.automations import EventTrigger, Posture, ResourceSpecification
from prefect.events.actions import CallWebhook

Automation(
    name="On deployment completion -> webhook filter by schedule",
    enabled=True,
    trigger=EventTrigger(
        expect={"prefect.flow-run.Completed"},
        posture=Posture.Reactive,
        match_related=ResourceSpecification({
            "prefect.resource.role": "deployment",
            "prefect.resource.name": "my-deployment"  # or match by id
        }),
        threshold=1,
        within=timedelta(seconds=0),
    ),
    actions=[
        CallWebhook(
            url="<https://your-service/handle-completion>",
            method="POST",
            payload={
                "flow_run_id": {
                    "__prefect_kind": "jinja",
                    "template": "{{ flow_run.id }}"
                }
            },
        )
    ],
).create()

Sketch of the webhook handler

from uuid import UUID
from fastapi import FastAPI, Request
from prefect.client.orchestration import get_client

TARGET_SCHEDULE_SLUG = "my-schedule-slug"
DOWNSTREAM_DEPLOYMENT_ID = UUID("bbbbbbbb-bbbb-bbbb-bbbb-bbbbbbbbbbbb")

app = FastAPI()

@app.post("/handle-completion")
async def handle(req: Request):
    data = await req.json()
    flow_run_id = data["flow_run_id"]

    async with get_client() as client:
        fr = await client.read_flow_run(flow_run_id)
        cb = fr.created_by
        if not cb or cb.type != "SCHEDULE":
            return {"skipped": "not from a schedule"}
        if (cb.display_value or "").lower() != TARGET_SCHEDULE_SLUG.lower():
            return {"skipped": "different schedule"}

        await client.create_flow_run_from_deployment(
            deployment_id=DOWNSTREAM_DEPLOYMENT_ID,
            parameters={"...": "..."}  # your params_a
        )
    return {"triggered": True}

Related docs - Automations overview: Automations - Chaining deployments with events: Chaining deployments with events - Using parameters in templates: Access parameters in templates If you can share whether separate deployments per schedule is acceptable, I can outline the exact automation config for your setup. If not, I can help you wire up the webhook approach.