[maya-ots]

What

Added the on_health_failure plugin hook so plugins can react when a container becomes unhealthy after deployment.
Closes #41

How

• Added a health monitor that periodically checks the health status of running instance containers.
• The hook is triggered when a container transitions to unhealthy.
• Avoided dispatching the same event repeatedly while the container remains unhealthy. If it recovers and becomes unhealthy again, the hook can be triggered again.
• Added the health failure payload with the instance, challenge, container, status, and timestamp.
• Started the monitor alongside the existing reaper in the instancer.
• Added tests covering unhealthy transitions, repeated checks, and recovery.

Verification

• go build ./...
• go test ./...
• golangci-lint run ./...
• mctf generate --check on the sample event (if derived files could change)
• Docs updated (docs/spec/, CHANGELOG) if behavior/CLI changed
• Schema + make schemas synced (if on-disk shape changed)

Notes

The health monitor uses Docker inspection to check container health, so containers without a Docker health check don't trigger the hook.
also, all tests related to our health-failure hook are passing :

• internal/instancer — PASS
• internal/plugin — PASS
• The other packages shown — PASS

but i got failures in internal/kubernetes. what appears to be failing :

1. Windows path separators: Tests expect pwn/pwn-me, but Windows produces pwn\pwn-me. The same issue affects web and misc.
2. Unexpected error behavior: TestE2E_ApplyUp_ErrorStopsExecution expected an error but didn't receive one. This may be a separate issue.

The Kubernetes test failures appear unrelated to this change and were left untouched and i got them too while working on a previous issue. i'll fix them soon as a new issue