All templates

Operations

Nightly data quality checks

Runs your invariants against production every night and tells the on-call analyst which rows broke and since when.

The prompt

This is the whole configuration

Paste it into a workflow and edit it in plain English. There is nothing else to wire up.

The prompt
Every night at 3am, run our data quality checks against the production Postgres replica. Check that: no order has a null currency, every account's plan is one of the four values in our enum, no invoice has a total below zero, every user row has a matching account, and the events table has received rows in each of the last 24 hours. For every failing check, find how many rows are affected, identify the earliest affected row and its timestamp, and look at what changed upstream around that time. Email the on-call analyst through Resend with only the failures, the affected row counts, the first occurrence, and the query you ran. When everything passes, send nothing.

Trigger

What starts this agent

Nightly at 3:00am

Runs on a schedule. Describe the cadence in plain English and Helios turns it into a cron expression for you.

Integrations required

Authorize each service once. Helios stores the credential envelope-encrypted and attaches it server-side on every run.

Example output

What you get back

Every run leaves a readable record: the steps the agent took and the result it produced.

2 checks failed

  • `orders.currency` null on 41 rows since 02:10
  • `accounts.plan` has 3 values outside the enum
  • 12 checks passed

New to Helios? Start with Run your first agent.

Run Nightly data quality checks on your stack.

Connect the integrations, paste the prompt, attach the trigger. The agent takes it from there.