API Analyst
services & ports
Overview
Server-side liveness of every SAGRIS service (ports 5001–5007), plus a live snapshot of the API host. The probe runs from the API process, so it reaches the localhost-bound services the browser cannot.
Service health
Host snapshot
request workbench
Endpoint tester
Send a real request to any of the 12 endpoints and inspect status, latency, payload size and response. Public endpoints need no key; the rest require a data-API key (X-API-Key) — paste yours below (kept only in this tab's sessionStorage). Edit the path/body inline before sending.
processing load
System load
Host CPU / memory / load average and per-endpoint request activity (latency sparklines, p50/p95, error counts) from the in-memory ring buffer shared across workers. The buffer holds the most recent requests and is lost on restart.
Host
Endpoint activity
Sparkline = latency of the most recent calls (low→high). Ordered by most-recent activity.
latency benchmark
Benchmark
Fire N requests at one endpoint with a chosen concurrency and report p50/p95/max and throughput — useful for spotting dc.load slowdowns. Default to the public /api/stats (cached) or /api/health to avoid heavy load; the data endpoints use your key and the body from the tester.