Overview

node-metrics provides development-only tooling for inspecting in-process Prometheus metrics. Production services still expose /metrics for Prometheus/Grafana; this suite parses that text format in memory and surfaces it through admin HTTP routes and the admin SPA.

Core runtime metrics collection lives in @saflib/node (collectSystemMetrics, subsystem histograms). This suite is the dev viewer layer on top.

What this suite provides

• node-metrics-spec — OpenAPI contract for the metrics snapshot admin route
• node-metrics-http — Express router and Prometheus text parser (parsePromText, createMetricsRouter)
• node-metrics-sdk — TanStack Query client hooks for the admin UI
• node-metrics-vue — Admin SPA page for browsing parsed metrics

Integration

When the repo follows base/service/http, the HTTP app mounts createMetricsRouter() in development. The admin SPA uses @saflib/node-metrics-vue to fetch /admin/metrics/snapshot and display metric families in a table.

No external observability stack or cloud keys are required for local development — metrics stay in the running Node process until scraped by this tooling or by production Prometheus.