Automate usage of Kai client to accomplish a migration scenario (no IDE in this flow)

[dymurray]

Goal here is to take coolstore as a first example and migrate it on the main branch to quarkus.

[abrugaro]

This workflow will:

1. Fetch the binaries from the latest kai release
2. Fetch the kai code and install all the dependencies.
3. Execute the run_demo.py script.
4. Clone the kai-eval repository from https://github.com/abrugaro/kai-eval
5. Parse the analysis logs and run the evaluation based on incidents reported by the analysis and files modified by kai
6. Generate a report based on the evaluation, including execution times, date and number of modified files and upload it to a s3 bucket that can be accessed by the visualizer.
7. Zip the coolstore repository, the kai and kai-eval logs and upload them to the same s3 bucket.

Notes:

1. The workflow is scheduled to run automatically once a day.
2. As of now this is only working for Linux, working now on the Windows version.
3. kai-eval was originally developed by Sam, however it was based on llm trace logs which are not stored anymore, so I modified it to be based on analysis output, and files diff. Due to that the visualizer will need a small adaptation to update some fields.

The workflow is failing on this repo right now, I believe it is due to the credentials, I'll check with Sajid.

Is this everything that was required for this task @dymurray? Can we close it once the workflow is green?

Edit: The pipeline is failing due to (I believe) an issue in the kai-analyzer-server when executing the run_demo.py script