Investigations to do during AI frontier week:

Using repo:  https://github.com/tesshuflower/aiexperiments/blob/main/fleet-mgmt-app-repos/

• [x] Use and train claude to manage PRs in repos to build volsync and fbc
  • [x] be able to identify konflux and tekton pipeline prs
  • [x] be able to rebase prs where renovate is not updating these in a timely fashion (tedious manual work normally)
• [x] Use and train claude to manage PRs in other business continuity repos
• [x] train claude to be able to find FBCs and images from PRs in order to be able to test on clusters.
  • [x] learn how to connect to various clusters, including konflux itself
  • [x] learn how to pull information from konflux resources (pipelineruns, snapshots) and pull them together.  This is normally very tedious when we have 7+ builds to find (one per OCP version, and this will increase to 8 with 4.20)
  • [x] learn how to create a catalogsource with PR fbc images on openshift clusters
  • [x] learn how to deploy the catalogsource
  • [x] learn how to deploy and monitor the operator
• [x] train claude on cluster management for our dev/test
  • [x] use the ck tool
  • [x] get the proper kubeconfig for a cluster
  • [x] determine if ImageContentSourcePolicy is setup (if required)
  • [x] update global pull secret with quay.io:443/acm-d creds (if required)
• [x] train claude to run the full e2e tests using custom-scorecard
  • [x] ensure the tests are run at the correct version
  • [x] notify when the tests are complete
  • [x] analyse the logs and report errors etc
• [] some additional konflux instructions
  • [x] how to auth with konflux cluster
  • [x] read the claude & cursor config from the konflux-release-data repo
  • [x] find snapshots with proper images for a volsync release - normally this is very tedious and perhaps error-prone
  • [] aid with creating a release yaml
• [] try out code reviews
  • [x] read and analyse large upstream PR
• [x] Setup Jira CLI with claude or gemini
• [x] possible to do a slack notification when I have claude run a long running task (for example running e2e tests)?