Logging and metrics

[nazargladish]

Problem

• The logger is difficult to use
• Debugging with it is challenging
• There's a lack of consistent logging across the codebase

Additionally, we have code snippets like this:

    if (this.debugAI) {
      console.log(pc.green(`Executing cached test ${hashData(test)}`));
    }

In my opinion, if(this.debugAI) should be an impl detail of the Logger service itself (DRY)

Metrics that would be valuable to collect (preferably in the cloud):

• Test execution time
• Input/Output tokens
• Execution cost

[rmarescu]

I've created an issue to improve logging, feel free to add any suggestions there.