Add worker observability metrics

[ianhodge]

Summary

• add bounded OpenTelemetry metrics for task assignment age, startup latency, failure classification, backend operations, cleanup failures, cancellations, and cancelled task results
• emit opt-in per-task traces and lifecycle events when OTEL_TRACES_EXPORTER is configured
• classify Docker, Kubernetes, and direct backend failures with stable phase and reason labels
• update README monitoring docs with tracing setup, the metric catalog, and sample dashboard queries

Validation

• go test ./internal/metrics ./internal/worker
• go test ./...
• helm template oz-agent-worker charts/oz-agent-worker --set image.tag=test --set worker.workerId=test-worker --set warp.apiKeySecret.create=true --set warp.apiKeySecret.value=dummy --set metrics.enabled=true
• git diff --check

Co-Authored-By: Oz oz-agent@warp.dev

[ianhodge]

• Add worker cancellation controls #63
• Add worker observability metrics #62 👈 (View in Graphite)
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[ianhodge]

Made these changes based on what figma asked for in https://docs.google.com/document/d/1gVD8eYVpjp7BRkoaV4_nfdObDDoHaqiuSfmjhYaVwns/edit?tab=t.0#heading=h.o3ms8i9t7i1h

[bnavetta]

Nice!

[bnavetta]

This one seemed useful?