Short answer: run prefect deploy from your project root. Prefect will read your prefect.yaml, run the build/push steps locally, and register the deployment(s). Then make sure a worker is polling your work pool. Here’s a step-by-step checklist for your template in Prefect 3.x: 1) Prereqs - Ensure you’re on Prefect 3.x (your file pins 3.4.18). - Install the Docker extra so the build/push steps are available:

pip install "prefect==3.4.18" "prefect-docker>=0.4.5"

Or run CLI commands with extras on demand:

uv run --with 'prefect[docker]' prefect deploy

2) Create the work pool (if it doesn’t exist yet) - If you’re running this in Docker containers, create a Docker pool:

prefect work-pool create fraio --type docker

- If you’re using Kubernetes instead, use --type kubernetes and make sure your job_variables match that template. 3) Start a worker for the pool

prefect worker start --pool fraio

Add --work-queue if you’re using a specific queue. 4) Deploy from the prefect.yaml - From the directory containing prefect.yaml:

uv run --with 'prefect[docker]' prefect deploy

Optionally target just one deployment in the file:

uv run --with 'prefect[docker]' prefect deploy --name fraio-trigger

5) Kick off a run (optional)

prefect deployment run fraio-trigger/local --watch

Important notes about your template - entrypoint must reference a valid Python function name. You have: ./fraio_algorithms/fraio_trigger_flow.py:trigger-flow That hyphen is not a valid Python identifier. It should be: ./fraio_algorithms/fraio_trigger_flow.py:trigger_flow - The image reference in job_variables looks right for Docker/Kubernetes pools: image: '{{ build_image.image }}' - Make sure you’re authenticated to your registry before deploying so the push step can succeed:

docker login harbor.mgmt.vintecc.cloud

Docs and references - Prefect 3.x deployments via prefect.yaml: Prefect YAML guide - Work pools CLI: prefect work-pool CLI - Workers CLI: prefect worker CLI - Docker collection (build/push steps): Prefect Docker collection Want me to validate your entrypoint and pool type are aligned with your runtime? If you confirm whether the ‘fraio’ pool is Docker or Kubernetes, I’ll tailor the job_variables and next steps accordingly.

Short answer: no. In Prefect 3.x the prefect.yaml cannot create a work pool; it only references an existing pool by name. You need to create the pool once via CLI, UI, or API before running prefect deploy. Typical ways to automate it before deploy: - CLI (Docker pool)

prefect work-pool create fraio --type docker --overwrite
prefect worker start --pool fraio
prefect deploy --name fraio-trigger

- CLI (Kubernetes pool)

uv run --with 'prefect[kubernetes]' prefect work-pool create fraio --type kubernetes --overwrite
prefect worker start --pool fraio
uv run --with 'prefect[docker]' prefect deploy --name fraio-trigger

You can put those commands in your CI/CD pipeline so the pool is created/updated before each deploy. Docs: - Work pools (CLI): prefect work-pool CLI - Deployments from prefect.yaml: Prefect YAML guide If you tell me whether your ‘fraio’ pool is Docker or Kubernetes, I’ll give you the exact create command and any job_variables you may want to set.

Yes — point

prefect deploy

at your file with the

--prefect-file

flag. Examples:

prefect deploy --prefect-file ./my_prefect.yml
prefect deploy --prefect-file ./deploy/prefect-deploy.yaml --name fraio-trigger

Tip: since your template uses Docker build/push steps, run with the Docker extra to ensure dependencies are present:

uv run --with 'prefect[docker]' prefect deploy --prefect-file ./deploy/prefect-deploy.yaml

Docs: Prefect YAML guide