[OBSERVABILITY]: Define SLOs and end-to-end tracing for payout operations

[barry01-hash]

Context

Console logs and product analytics do not answer whether payout creation, application approval, escrow funding, or milestone release is healthy. Operators need correlated telemetry across browser, API, database, jobs, and blockchain dependencies.

Scope

• Define SLIs/SLOs for availability, latency, and successful completion of critical workflows.
• Add structured logs, metrics, and distributed traces with shared correlation/operation IDs.
• Instrument API latency/status, database timing, job queues, escrow confirmation latency, and dependency failures.
• Redact tokens, emails, wallet-sensitive metadata, request bodies, and file URLs by default.
• Build dashboards and burn-rate alerts tied to actionable runbooks.
• Add synthetic probes for public availability and authenticated test-environment journeys.

Acceptance criteria

• A single payout/escrow operation can be traced across all participating components.
• SLO calculations exclude only documented cases and use version-controlled definitions.
• Alerts use multi-window burn rates and link to recovery runbooks.
• Telemetry has explicit sampling, retention, and cardinality limits.
• Automated tests verify redaction of secrets and PII.
• This work integrates with, rather than duplicates, the functional system-log issues [FEAT]: Implement System Logs Routes #116 and [FEAT]: Integrate System Logs Across Application #117.

Relevant areas

API routes, Prisma client, external adapters, escrow hooks/workers, PostHog provider, deployment platform.