Failed deploys say who needs to act and how to fix it | Ample changelog | Ample

Deployment Failures

Failed Deploy: A failed deploy now indicates who is responsible (your app, your account, or Ample), including a diagnosis, log evidence, and a suggested fix, such as a missing environment variable.
Ample Failures: When a failure is on Ample's side, the deploy waits while Ample fixes it and then continues. If it still fails, the error links to an incident.
CLI Command: ample auth claim --email (MCP: ample_auth_claim) sends an email to the person with a link to claim the account. The sign-in page also offers to claim an unclaimed account.

Budget Management: Apps that exceed their budget are now paused, with their data retained and not deleted. They respond with a 503 budget_exceeded message until the account is back under budget.
CLI Plan Detection: ample plan detects standard-library Python apps and outlines an empty plan. Rerunning ample deploy while a deploy is in progress will wait for it instead of failing.
File Management: Files located under node_modules, vendor, .venv, target, and similar folders no longer determine an app's runtime.
Start Script Preservation: An app's own start.sh will not be overwritten. Each runtime with an install or build step receives an 8 GiB disk, and refused boots will now report the reason.