2026-09-23
Scheduled drift detection now runs for every customer
#atlantis #tofu #terragrunt #automation
Last month we piloted automated drift detection with a couple of customers. It now runs for every managed repository: a rolling report that surfaces infrastructure changed outside your normal pull-request workflow.
How it works
- A scheduled job asks the Atlantis instance to plan every stack in your repository against the main branch.
- The result lands as a single, rolling “Terraform drift report” issue in your repository: which stacks are clean, which have drifted, and which couldn’t be checked (for example because a plan is already running from an open pull request).
- Only a summary is posted, no full plan output. Remediation stays a human decision: you or we follow up with a regular pull request on anything that drifted.
What you need to do
Nothing. This is switched on for every customer, no configuration required.