Editorial Policy

Last updated: 2026-10-10 · Version: v1

How reviews get published

Every review on RepoVerifier, whoever wrote it, goes through the same process:

1. Tested. A person — the site's founder, or a community member — actually runs the repository against a claim it makes, and records what happened.
2. Checked. Automated checks look at basic facts: does the repository exist and is it public, and is the claim consistent with the repository's own documentation.
3. Reviewed and approved. An editor reviews the submission — including, for the most serious verdicts, independently attempting to reproduce the finding — before anything is published. Nothing is published automatically.

Two tracks, one bar

• Founder-tested reviews are written and tested by RepoVerifier's founder.
• Community-tested reviews are submitted by someone else who tested the repository themselves, and are reviewed and approved the same way — same evidence standard, same editorial judgment, same final sign-off.

Both tracks are labeled clearly on every review, listing, and certificate. Neither is presented as more or less legitimate than the other; the label describes who did the testing, not how carefully it was reviewed.

What "reviewed" means — and doesn't

We describe review content precisely, and avoid words that overstate what happened:

• We say a repository was tested and a review was reviewed, not that anything was formally "verified" in a technical sense.
• Certificates are titled "Testing Certificate" (positive verdicts) or "Testing Record" (Overhyped/Dangerous verdicts), never wording that implies an endorsement or a formal verification process.
• Evidence is human-reviewed, not independently execution-verified. An editor reads and evaluates the evidence a submitter provides. RepoVerifier does not claim to have independently re-executed every test in every submission.
• Reviewer handles are randomly generated and are not identity-verified. We do not confirm that a handle or display name belongs to a specific real person or organization.

Payment and independence

Submitting and publishing a review is free. Payment, sponsorship, or any other consideration can never influence whether a submission is approved or what verdict it receives. We do not run pay-to-submit or pay-to-publish programs.

Affiliation disclosure

If you are the author or maintainer of a repository, work for or are paid by the project or its company, or otherwise have a connection to it, you must disclose that when submitting a review of it. We show this disclosure to the editor alongside your submission. An undisclosed affiliation, if later discovered, can result in removal of the review.

We are transparent about a real limitation here: we cannot technically verify whether a submitter is affiliated with a project. This policy, and editorial judgment at review time, are the actual safeguard — not an automated check.

Verdicts

Every review reaches one of five verdicts: Hidden Gem, Solid, Partial, Overhyped, or Dangerous. Overhyped and Dangerous verdicts make a public, negative statement about a real project, so they are held to a higher bar: stronger evidence is required (a reproducible fork or branch is preferred over a screenshot), and an editor independently attempts to reproduce the finding before approving. If we cannot reproduce a serious negative finding, we do not publish it as submitted.

What to submit

A public GitHub repository URL, the claim you tested, a verdict (Hidden Gem, Solid, Partial, Overhyped or Dangerous), your testing notes (what you ran, where, and what happened), a score from 0 to 5 on each of the five criteria, and your connection to the project (none, author, employee or other; if other, please explain). A fork link and evidence links are optional, but we expect a fork or branch for Overhyped and Dangerous verdicts.

What we don't publish

Reviews of repositories you haven't run; copied, untested or fabricated reviews; personal attacks, hate, harassment or anything illegal; secrets or other people's personal information; and reviews written to promote or damage a project for payment or competitive gain.

Outcomes

A submission is approved, returned for a fix (including when we could not reproduce your result), or rejected. Fixes and rejections carry a reason, shown on your status page and in the email. For a final rejection the email does not repeat the reason, but the status page shows it.

After publication

Only the editor can take a review down. The author is told it was withdrawn, without a reason. Project maintainers can use the report link on the certificate page, or the contact page.

Age

You must be 16 or older to submit a review.

Reporting a review

Anyone — including a project's own maintainer or contributor — can report a published review, using the report link on the certificate page, or the general contact page. Tell us what you believe is wrong and why. We aim to respond within a week.

If a report is substantiated, we remove the review, revoke its certificate, and keep a private record of the decision. We do not disclose who reported a review to the person whose review was reported. Removing a review is not automatically reversible; if you believe a removal was a mistake, use the same contact process.

Founding tester recognition

The first 25 approved community reviewers receive a "Founding tester" designation on their public profile. This recognizes early participation in reviewing; it is not a rank, a credential, or a claim about the quality of any individual's reviews.

Questions or disputes

Editorial decisions (approve, reject, request changes) are made by RepoVerifier's editor and are final unless a factual error in the decision is shown. If you think a decision was made in error, or you have any other question about this policy, use the contact page.

This is a plain-language policy for an early-stage project and has not been reviewed by a lawyer.