ReleaseProof — visual evidence before you ship

ReleaseProof compares authorized public preview and production pages across mobile, tablet, and desktop. It produces responsive screenshots, red-highlight visual differences, browser findings, and a shareable QA report your team or client can review without installing a test framework.

Open a real ReleaseProof example report

Why teams use ReleaseProof

• Review a release before launch. See the candidate page at three practical viewport sizes.
• Compare preview with production. Match every candidate path and query to one production origin.
• Catch observable problems. Find horizontal overflow, failed assets, browser errors, blocked pages, timeouts, and unexpectedly tall screenshots.
• Share proof, not a vague score. Send one read-only HTML report with the screenshots, findings, visual diffs, JSON, and CSV.
• Keep failures honest. A navigation or processing failure is reported as failed, blocked, timed out, or unknown—never silently converted into a pass.

ReleaseProof is useful for agency handoffs, landing-page launches, redesigns, CMS changes, e-commerce releases, and regression checks where a visual record matters.

What one audit includes

For every submitted page, ReleaseProof can provide:

• full-page PNG screenshots at mobile (390×844), tablet (768×1024), and desktop (1440×900);
• isolated candidate and production-baseline captures for each selected device;
• a changed-pixel percentage and red-highlight diff image when comparison succeeds;
• final URL and main-document HTTP status;
• console errors, uncaught page errors, and failed image, font, stylesheet, script, XHR, or fetch requests;
• horizontal-overflow evidence with up to 20 visible offending elements;
• page dimensions and clear screenshot-truncation disclosure;
• one ordered dataset row plus matching HTML, JSON, and CSV outputs.

URL order and duplicates are preserved. One problem page does not cancel unrelated pages.

How it works

1. Add 1–20 candidate URLs. These must be public pages you own or are authorized to assess.
2. Optionally add the production origin. ReleaseProof appends each candidate path and query to that origin. For example, https://preview.example.com/pricing?currency=usd is compared with https://example.com/pricing?currency=usd.
3. Choose the devices and run. Open the HTML report from the Output tab and share its hard-to-guess link with the people reviewing the release.

{
  "urls": [
    "https://preview.example.com/pricing?currency=usd"
  ],
  "baselineBaseUrl": "https://example.com",
  "devices": ["mobile", "tablet", "desktop"],
  "navigationTimeoutSeconds": 45,
  "settleTimeMilliseconds": 1500,
  "confirmAuthorization": true
}

Result statuses

Status	Meaning
passed	Every requested capture completed with no reportable findings.
warning	Capture succeeded with browser errors, failed resources, truncation, or at least 1% visual difference.
failed	The main document returned another HTTP error, or horizontal overflow exceeded 2 CSS pixels.
blocked	The site returned 401/403/407/429/451, showed a recognized challenge, or attempted an unsafe redirect.
timed_out	Navigation or capture exceeded its configured deadline.
unknown	DNS, browser, capture, comparison, or another indeterminate failure prevented a reliable conclusion.

Visual difference is evidence, not automatic proof of a defect. A person should review the images.

Transparent pricing

• $0.05 per submitted page audit. Every selected device, optional production comparison, screenshots, available diff images, findings, HTML report, JSON, and CSV are included.
• $0.00005 Actor-start event per GB of configured memory. The default 2 GB run therefore adds $0.00010 once per run.
• Platform usage is included. Customers do not receive a separate infrastructure charge.

One URL is one page audit. At the default memory, 1 page costs about $0.05010, 10 pages about $0.50010, and 20 pages about $1.00010.

The page event is charged only after the report, exports, dataset row, screenshots, and available diffs have been stored successfully. A failed, blocked, timed_out, or unknown page remains a completed audit when ReleaseProof delivered the promised evidence. Invalid input does not produce a page-audit charge.

You can set a maximum run charge in Apify. ReleaseProof checks that it covers all submitted pages before opening Chromium.

Trust, safety, and privacy

ReleaseProof is deliberately narrow and evidence-based:

• it accepts public HTTP(S) pages only—no credentials, cookies, custom headers, proxies, CAPTCHA solving, or login automation;
• it blocks localhost, private, link-local, reserved, cloud-metadata, credential-bearing, non-HTTP, and browser-internal URLs on initial requests, redirects, and subrequests;
• candidate and baseline pages run in isolated browser contexts;
• diagnostic resource URLs are stored without query strings, and message counts are bounded;
• reports contain no ReleaseProof tracking scripts and are rendered from autoescaped data;
• no AI model assigns design scores or invents findings.

Run artifacts are intentionally readable by anyone possessing their hard-to-guess storage URL and remain available according to the customer's Apify storage-retention policy. Do not submit private or secret information through public page URLs.

Honest limits

• Screenshots are capped at 16,000 CSS pixels and explicitly marked when truncated.
• Dynamic content, consent dialogs, ads, dates, live inventory, and experiments can create legitimate visual differences.
• ReleaseProof observes the loaded page; it does not click through flows or validate business logic.
• Scheduled monitoring, Lighthouse, accessibility/SEO audits, PDF reports, custom branding, authentication, and AI design criticism are not included in this version.

ReleaseProof reports observable browser evidence, not a guarantee of correctness, accessibility, security, or legal compliance. Test only pages you own or are authorized to assess, and review the evidence before making release decisions.