I’m not sure what you want me to try again. Could you share a bit more detail about what you’re working on? Helpful details: - What are you trying to do (and with which Prefect version—assuming 3.x unless you say otherwise)? - How are you running it (Cloud or self-hosted, worker/work pool type)? - The exact command/code you ran - The error/log output you’re seeing If you want me to re-run or re-share specific steps or code from earlier, just paste the last thing you want me to “try again” and I’ll jump on it.

Short answer: Helm configures the Prefect worker pod, but your DB env vars need to go on the flow-run pods. In Prefect 3.x you set those on the Kubernetes work pool’s base job template (or per-deployment via job_variables). Helm can create your Kubernetes Secret, but it cannot inject env vars into flow-run pods by itself. What to do - If these env vars should apply to all flows in the pool: add them to the work pool’s base job template. - If they’re per-deployment: pass them as job_variables in your deployment. Key distinction - Worker pod (managed by Helm): long-lived, polls for work. Env here does NOT reach your flows. - Flow-run pod (created per run by the worker): where your code runs; put DB env here via the work pool base job template or deployment job_variables. Option A: Use a Kubernetes Secret and reference it in the base job template (recommended) 1) Create a Secret in your namespace (can be via your own Helm chart):

apiVersion: v1
kind: Secret
metadata:
  name: my-db-secret
type: Opaque
stringData:
  DB_USER: myuser
  DB_PASSWORD: mypass
  DB_HOST: myhost
  DB_NAME: mydb

2) In Prefect UI → Work Pools → your Kubernetes pool → Advanced, ensure the base job template’s first container includes both envFrom (for the secret) and the existing env placeholder. In the job_manifest under spec.template.spec.containers[0], add:

"envFrom": [
  { "secretRef": { "name": "my-db-secret" } }
],
"env": "{{ env }}"

This makes all keys in the secret available as env vars on every flow-run pod, and still lets you add/override vars via job_variables.env. Option B: Set per-deployment env with job_variables In code:

from prefect import flow

@flow
def my_flow():
    import os
    print(os.getenv("DATABASE_URL"))

if __name__ == "__main__":
    my_flow.deploy(
        name="my-deployment",
        work_pool_name="my-kubernetes-pool",
        job_variables={
            "env": {
                "DATABASE_URL": "<postgresql://user:pass@host:5432/dbname>"
            }
        }
    )

This populates the

{{ env }}

placeholder in the base job template for this deployment. Option C: Reference individual keys from a Secret in env If you prefer explicit env entries instead of envFrom:

"env": [
  {
    "name": "DB_USER",
    "valueFrom": { "secretKeyRef": { "name": "my-db-secret", "key": "DB_USER" } }
  },
  {
    "name": "DB_PASSWORD",
    "valueFrom": { "secretKeyRef": { "name": "my-db-secret", "key": "DB_PASSWORD" } }
  }
]

Notes - You cannot set flow-run env vars via the Prefect Worker Helm chart; that chart only configures the worker pod. Use it to deploy the worker and optionally to create your Secrets; then reference those Secrets in the work pool base job template as shown. - The default Kubernetes work pool template already includes

"env": "{{ env }}"

on the container; keep that so you can pass overrides via job_variables. - For security, prefer Secrets over putting DB creds inline. Helpful docs - Customize base job templates: Customize base job templates - Customize job variables per deployment: Customize job variables - Kubernetes integration overview: Prefect Kubernetes integration - Prefect Helm charts: prefect-helm Quick clarification to tailor this: - Do you want these DB env vars applied to all flows in the pool, or only certain deployments? - Do you already have the DB credentials in a Kubernetes Secret (and what’s the namespace/secret name), or should we show a quick Helm snippet to create it?