Maintenance Agent Test

A sample Ruby application demonstrating various services.

Logging Approach

This project currently uses simple puts statements and direct Logger.new(STDOUT) calls for logging throughout the codebase. Each service handles its own logging in whatever way the original author preferred:

• API server/middleware: Mix of puts and Logger.new(STDOUT)
• Worker: puts and $stdout.write for job processing output
• Mailer: puts and Logger.new(STDOUT) for email operations
• Scheduler: Mix of puts and Logger.new(STDOUT) in health checks
• Importer: puts and $stderr.puts for import progress and validation errors
• Scripts: puts for progress output

Why This Works (For Now)

For local development, puts gives us quick visibility into what's happening. In production, stdout is captured by the process manager, so we still get basic output. It's simple and everyone understands it.

Known Limitations

• No log levels (everything is effectively "info" or just raw output)
• No structured formatting (timestamp, severity, caller info)
• No way to filter noisy output in production
• Inconsistent formatting across services
• $stderr.puts used ad-hoc for errors instead of a proper error-level log

Note

There is a standardized logger in lib/app_logger.rb that was set up a while back, but it hasn't been adopted by any of the services yet.