@tensorfeed/status-widget

Drop the live TensorFeed AI status monitor onto any site with one line. Real-time operational status and p95 latency for every major AI provider and service, in a self-contained sci-fi console. Free, no API key, no tracking, zero dependencies.

Live preview and docs: https://tensorfeed.ai/embed

Install

npm install @tensorfeed/status-widget

Use it (any framework, or none)

import '@tensorfeed/status-widget';

<tensorfeed-status accent="blue" poll="30" height="600"></tensorfeed-status>

The custom element renders the widget in a shadow root, so host-page CSS cannot affect it and it cannot affect your page.

React

No custom element needed. Use the URL helper:

import { tensorfeedStatusSrc } from '@tensorfeed/status-widget';

<iframe
  src={tensorfeedStatusSrc({ accent: 'blue' })}
  title="TensorFeed live AI status"
  loading="lazy"
  style={{ width: '100%', maxWidth: 720, height: 600, border: 0 }}
/>

Plain HTML, no build

<script type="module">
  import 'https://esm.sh/@tensorfeed/status-widget';
</script>
<tensorfeed-status></tensorfeed-status>

Attributes

| Attribute | Values | Default | Notes | |---|---|---|---| | accent | blue | auto | green | blue | blue is a light-blue spine with green status indicators. auto greens the whole accent when all systems are nominal. green forces green. | | poll | 5 to 600 | 30 | Client poll interval in seconds. Raise it on low-traffic pages. | | height | px number or any CSS length | 600 | Widget caps at 720px wide and reflows down to ~320px. | | label | string | TensorFeed live AI status | Accessible iframe title. |

What it shows

Operational status and probed p95 latency where TensorFeed measures it, real 7-day uptime % otherwise. Vendor status is authoritative; a provider with no status source is shown as "no data", never a false alarm. "Detail" deep-links to the per-provider page on tensorfeed.ai.

License

MIT. The widget data comes from the public TensorFeed endpoints /api/status/summary and /api/status/leaderboard.