slopwash

Terms

What the scores mean

Slopwash scores measure how densely a text matches the patterns in a published standard for AI-assisted writing. A score does not determine whether a human or an AI wrote the text, and it is not a judgment of the writing as a whole. Findings are editorial input. The decision to publish stays with the editor.

The free tools

The scanner, the demo, the prompt, and the public MCP endpoint are provided as-is, with rate limits, and they may change. The ruleset and the analyzer carry version numbers, and a result names the versions that produced it.

Audits and pilots

An audit is a written report on a document set agreed in advance, delivered at the stated price. Findings name passages and patterns; an audit is not a guarantee of any outcome with readers, platforms, or detectors. Team pilots are set up directly, billed monthly, and can be ended by either side at the end of a billing month.

API keys

Keys are issued to a named holder, carry the tier and scopes set at issue, and can be revoked for use that breaks the stated rate limits or for resale of access. Draft and rewrite text from a keyed call is kept only when the key has opted in to evaluation storage; by default nothing is kept. Data handling for keyed calls is described on the data handling page.

Questions go to @slopwash on X.