Editorial and testing methodology

Who maintains Anvil Tools

Anvil Tools is operated and maintained by VelloxTech. Guides are attributed to the VelloxTech Editorial Team because the work is maintained collectively; we do not invent individual biographies, credentials, or independent reviewers.

Report an error or suggest a correction. Include the page URL, the statement or behavior at issue, and a small synthetic example. Never send passwords, live tokens, private messages, or confidential documents.

How tools are tested

1. Define a real question and a fixed expected result before the run.
2. Create harmless synthetic inputs without personal information.
3. Run the public interface, not a separate hidden implementation.
4. Save the observed result, environment, date, screenshot or output, and downloadable fixture.
5. Use an independent check when practical—for example, reopen a generated PDF with a parser or compare an exact hash.
6. Publish limitations and failed cases alongside successful results.

The Anvil Tools Lab contains the reproducible reports and fixture hashes.

Browsers, devices, and coverage

The current recorded browser suite includes Chrome 154.0.8037.97 and Edge 154.0.4258.53 on a 1100 × 850 desktop viewport. Responsive layout checks use additional automated viewport sizes. A report names physical-device or printer testing only when it actually occurred; automated Chromium checks do not establish compatibility with every browser, phone, assistive technology, printer, email client, or external service.

Inputs, expected outputs, and privacy

Tests use invented text, generated files, public-format examples, or attributed licensed media. Fixture pages label synthetic data and publish SHA-256 hashes so visitors can verify exact bytes. Sensitive production data is not required for a useful reproduction.

A passing fixture proves the recorded case in the stated environment. It does not prove unlimited file capacity, regulatory compliance, perfect accessibility, or correctness for every input.

How sources are selected

Technical claims prioritize primary sources: specifications, standards bodies, browser documentation, official product documentation, and direct measured results. Secondary sources may provide context but do not replace the controlling specification or an observed test. Links are reviewed for relevance and claims are phrased within the evidence available.

AI assistance and fact-checking

AI may assist with drafting, outlining, editing, test-case suggestions, or identifying statements that need verification. Assisted material is not published automatically. VelloxTech manually reviews the final text, checks factual claims against cited sources or recorded results, removes unsupported claims, and keeps safety-sensitive examples synthetic. AI assistance is not presented as an independent reviewer.

Review dates, updates, and corrections

A visible “Last reviewed” date is shown only while the guide title, excerpt, and body match the reviewed version recorded in the repository. Material changes invalidate that marker until the new version is reviewed. Corrections preserve the previous database version and are recorded in an audit log when the publishing workflow supports it.

Small typographic repairs may not change the conclusion. Factual, safety, privacy, or test-result corrections should be described in the relevant report’s change history.

Advertising separation

Advertising is not a substitute for useful publisher content. Private temporary-email controls, inboxes, message readers, generation controls, copy/download workflows, and pages without meaningful editorial content must remain free of advertising placements.