The same bar applies to everyone

This library has one defence against becoming another shelf of confident advice: a protocol cannot be published without an account of its evidence and a statement of what it does not claim. That check is mechanical — it runs at build time and it fails the deployment, not a review meeting.

Contributions clear the same bar. That is the whole reason this page can exist without damaging anything: a decline points at a threshold rather than at a judgment, and the threshold is published.

How it works

1. You submit. No account, no email, no name. You get a key.
2. Triage. Checked for spam and abuse only — not judged. Anonymous text is not published automatically, because that would make this site an open publishing platform for strangers.
3. The open queue. Once through triage it is public, with its status. From this point nothing happens quietly.
4. Accepted or declined, in public, with the reason. A decline names what was missing. If the reason is wrong, that is itself a correction you can send.

Offers of work are the exception: they are not published, because they often carry contact details. They move through the same states, visible to you through your key.

What gets declined

• A proposal with no citations. It cannot be assessed, so it is not.
• A proposal with no statement of limits. Every published protocol carries one; there is no version of this library where contributions are exempt.
• A procedure too vague to run. If nobody could execute it on paper without you, it is not a protocol yet.
• Anything whose appeal is that it flatters the reader.

The most valuable submission is not a new protocol. It is a correction — a misread citation, an evidence rating more generous than it should be, a limit that should have been stated and was not. Those are acted on first.

Read the published protocols and one of their evidence files before proposing. It is the fastest way to see where the bar actually sits. What the server stores from this form is set out on the security page.