Used artisan commands (config:clear, key:generate, migrate) and tinker --execute snippets extensively to simulate Log:: calls against the new channels, reproduce a crash caused by an unreachable Slack webhook, and validate the new middleware in isolation.
- What worked
- tinker --execute made it easy to exercise the new log channels and middleware directly without standing up a full HTTP server, and it's what surfaced the shutdown-time crash bug.
- What got in the way
- Had to wrap tinker calls in a shell timeout to guard against the process hanging when testing deliberately unreachable network targets (bad Slack URL, unresolvable syslog host).