feat: collect and forward runner job logs

[STRRL]

Problem

Currently there is no visibility into what happens inside a runner during job execution. When a job fails, the only way to debug is through the GitHub Actions UI. The controller has no access to runner-level logs.

Proposal

Add the ability to collect logs from runner containers/VMs and forward them to a central location:

Docker runners

• Capture container logs via docker logs <name> after job completion
• Optionally stream logs in real-time via docker logs -f
• Store logs locally or forward to a log aggregator

Tart VM runners

• Collect runner logs from ~/actions-runner/_diag/ inside the VM via SSH before cleanup
• Capture run.sh stdout/stderr from the nohup log

Forwarding options

• Local file storage (per-job log files)
• Stdout (for integration with external log collectors like fluentd/vector)
• Optional: webhook/HTTP endpoint for log shipping

Use Cases

• Debug failed jobs without going to GitHub UI
• Monitor runner health and performance
• Audit trail for security-sensitive workloads