If your team manages customer-facing SaaS integrations through a vendor’s web UI, you already have configuration drift. Someone changed an OAuth scope at 11 PM to fix a stuck connection, a customer success manager toggled a field mapping for one enterprise account, or an engineer rotated a webhook URL during a postmortem and forgot to update the staging environment. When a third-party API sync fails at 2 AM, the root cause is rarely a complete vendor outage. It is almost always a configuration mismatch. None of these manual tweaks are in Git. None of them survive a rebuild.