In this video· click a step to jump to it
Test recording
APITube — app walkthrough (authed)
109,029 ms · 416,715 pixels changed
We built this regression test for dashboard.apitube.io — it's yours to keep, free.
Take this test with you
This is a real Playwright test, recorded against dashboard.apitube.io. Claim it into your own workspace and re-run it on every deploy — free.
export async function test(page, baseUrl, screenshotPath, stepLogger) {
const APP = 'https://dashboard.apitube.io';
const WS = '/ws_Z2MmQptgultfyocfXcDpvE';
const PLAYGROUND = WS + '/playground?endpoint=%2Fv1%2Fnews%2Feverything&is_breaking=1&language_code=en&sort_by=published_at&per_page=10';
let __step = 0;
const shot = async (label) => {
__step++;
const tag = String(__step).padStart(2, '0') + '-' + label;
const p = screenshotPath.replace(/(\.png)?$/, '-' + tag + '.png');
await page.screenshot({ path: p, fullPage: true });
stepLogger.log('📸 ' + __step + ': ' + label);Checks run
Notes from the demo run
AI-generatedWe took your 'no call, no form, no sales rep' claim literally: signed in with Google, landed on Account home with the key already masked and a quickstart pre-filled, opened the playground on the Breaking-news preset and executed a real request — 200 OK, 897 ms, 10 live articles — then watched that call show up in API logs with p50/p95/p99 within seconds. Then Explore data, Datasets, Integrations, Webhooks, API keys, Audit log, Subscription and Profile. The claim holds: first real response took under a minute from landing.
- The promise in the tweet is trueSign in, key is already there, playground preset, Execute search, 200 OK with 10 dated articles. No demo call, no key request form. That is genuinely rare in the news-data category and it is your strongest differentiator.
- Key masked by defaultapi_live_t7R…UICH renders masked everywhere including the code snippets, so a screen-share or a recorded walkthrough cannot leak it. Most API dashboards get this wrong.
- The payload is the productEach article came back with per-sentence sentiment, Flesch-Kincaid readability, target audience, keywords and share counts. The Visual tab makes that legible without reading JSON, which is the right call for the evaluate-in-5-minutes visitor.
- Logs close the loopRequests, billed requests, error rate, avg latency and p50/p95/p99 appeared immediately after our call. Being able to see your own request land is a quiet trust-builder.
- The profiling survey ambushes a deep linkThe 4-step 'What best describes you?' modal fired while we were on /logs, not on first landing, and exposed no Skip or close control — you must answer all four steps to get back to the page you were already on. It covered 9 of our 14 screens on the first pass. Firing it on the home screen, or adding a Skip, would keep it from interrupting work in progress.
- Playground runs spend the free allowanceThe header counted down 100 → 98 → 97 as we experimented. A visitor who tries three filter combinations has spent 3% of their free budget before writing a line of code. Either make playground runs free or put 'uses 1 request' on the button rather than in the sidebar.
- API key creation / reveal·Deliberately never revealed or minted a key — a real credential must never reach a public share
14 visual changes
Share this run
Proof your app is not AI slop
This test was built for dashboard.apitube.io — claim it free
One click copies the test and baseline screenshots into your own Lastest workspace. Re-run on every deploy and catch regressions before your users do — free, no card required.