The tool auto-checks the full catalog the moment it opens — 96 services across telecom, social, streaming, food, shopping, banking, travel, government, cloud/dev and gaming, Indian services first in each category.
Narrow the view with a category chip (for example Food & Grocery) and the region toggle (India or Worldwide), then press Check now to re-probe just what is on screen.
Every card shows a status pill — Up (green), Degraded (amber), Down (red) — plus the HTTP status code and the response time in milliseconds. The summary line above the grid totals the results and stamps the time of the last check.
Filters, results and the last-checked time persist in local storage, so a re-visit during an outage opens on the same view with the previous statuses still visible while the refresh runs.
What the statuses actually mean
Each probe is a HEAD request with a three-second timeout that follows no redirects. Up means the host answered with anything below 500 — including 403 and 405, which many sites deliberately return to datacenter requests; a response is still a response, and treating it as an outage would produce false alarms. Degraded means a 5xx answer: reachable, but the origin is erroring, which is the classic partially-down state. Down means nothing came back at all — timeout, DNS failure or refused connection, which are indistinguishable to the caller and all count as unreachable.
Reading a check honestly
The probe runs from DevBench's servers, so it answers "is this service reachable from the internet?" — not "is my connection working?". If every card is green but one site fails for you, look local first: your ISP, DNS resolver, VPN or the site blocking your network. If a service flips between Degraded and Up across a few checks, an incident is likely in progress — re-check in a minute to watch it settle. Results cache in this browser only, and the request allowlist is fixed: the tool probes exactly these well-known public hostnames and nothing else.