curriculum-project-hub/.scratch/saas-production-readiness/issues/30-instrument-correlated-runtime-telemetry.md

Instrument correlated production runtime telemetry

Type: task Status: open Blocked by: 13, 23, 24, 25, 27, 28, 29

Question

Implement the decided telemetry control plane with one redacted structured context from HTTP/Feishu ingress through durable work, run/session/lock, provider and SDK retry/rate-limit/tool events, cph/workspace operations, outbound delivery, audit, startup/drain/recovery, and release/schema identity. Expose bounded-cardinality metrics, dashboards and alert rules; eliminate silent critical-path failures; and prove logs can pivot end-to-end while metric labels contain no tenant/user/run cardinality or secret/message content.