Has anyone seen this error before?

* error configuring Terraform AWS Provider: failed to get shared config profile, dp_dev

This started happening 4 days ago. I’m a bit baffled because re-running a previously succeeded pipeline doesn’t get this error, yet the (very minimal) code changes at the time that the error started don’t seem to be related to the error at all.

Do you maybe have a custom config file set for AWS? Try

env | grep AWS

but of course PLEASE DO NOT POST ANY SECRET VALUES. (Obnoxious caps for safety.)

There’s no AWS config file, AWS creds are being provided by our build system through environment variables. These haven’t changed and aren’t different between builds that have this error and builds that succeed.