architect-monitoring

Monitor Cloudflare Worker health by collecting observability evidence from PostHog and Cloudflare logs, then mapping those signals to task-level actions to detect worker failures and error regressions during deploy or test workflows.

Slack escalation works by posting to the #error-pulse channel when incident impact is broad or unclear, attaching deep links to PostHog, Cloudflare log queries, and GitHub permalinks to provide evidence-based triage context for fast resolution.

Update the active task with monitoring evidence only when the incident situation is blocked, deferred, ambiguous, or repeated, ensuring that task records remain focused on evidence-only findings rather than routine status checks.

Evidence-based triage for OS and worker health incidents is the process of collecting observability signals, determining what changed or broke, and attaching direct links to PostHog and Cloudflare logs to operationalize fast incident response.

Handle missing credentials by prioritizing environment variables first, and use the `doppler run` command to inject required secrets when credentials are missing during the worker health monitoring and triage workflow.

The best way to triage repeated or ambiguous incident signals across environments is to pull events and traces to determine what changed, record evidence-only updates on the active task, and escalate to Slack when impact remains unclear.