The agents that need to rerun a build to verify their work (build repair) should be able to trigger it at the end of their job. Initially, we can present a running build to a developer. Later on, when supported, we can resume the agent's session if CI fails.
The build repair agent can run only a build for a specific platform where it failed, instead of the full CI.
The agents that need to rerun a build to verify their work (build repair) should be able to trigger it at the end of their job. Initially, we can present a running build to a developer. Later on, when supported, we can resume the agent's session if CI fails.
The build repair agent can run only a build for a specific platform where it failed, instead of the full CI.