Remediation

Knowing which resources are wasteful is only useful if fixing them is easy. For supported finding types, Parsivex generates the exact commands that clean up the specific resource it flagged — the volume ID, the instance ID, the bucket and prefix already filled in — so you are not translating a report into shell commands by hand at the end of a long day.

These scripts are deterministic templates built from the finding data, not text an AI made up on the spot, and they are the same for everyone who hits the same finding type. Parsivex never runs them for you. You copy them into your own terminal or pipeline, with your own credentials and your own change process, which means the decision to delete anything stays where it belongs. The read-only IAM role you granted for scanning cannot make these changes even if you wanted it to.

The articles here explain which findings come with a script and where to find it in the app, when to take the AWS CLI version versus the Terraform snippet, and how to run either one safely — reviewing the warnings attached to a script, using dry-run flags where they exist, and testing in a non-production account first. Read the safety article before your first cleanup, particularly for storage lifecycle changes, where the effect is not always immediately reversible. Remediation scripts are also deliberately private: they never appear in a shared public report.