Platform policy

Commentary Policy

How Author's Commentary sections are written, verified, and published on Sphinnx.

The sitewide commitment

Every Author's Commentary reply on Sphinnx is typed directly by the author without AI assistance or copy-paste functions. All replies are published verbatim and in full. No editing, no paraphrasing, no omissions.

What Author's Commentary is

Author's Commentary sections are structured Q&A blocks embedded in each article. The questions are generated by an AI reading the article — anchored to specific sections where the author's direct voice adds something the text alone cannot provide. The author then types their replies.

These are not summaries, not paraphrases, and not AI-generated answers. They are the author's own words, in response to a specific question, placed inline where they are most relevant to the reader.

How the input is controlled

The author types their answers in a plain text interface with paste disabled at the browser level. There is no AI completion, no suggestion engine, no clipboard access. The only input mechanism is the keyboard.

Each answer carries a minimum of 20 words and a maximum of 300 words. The system records the timestamp of first keystroke and submission for each answer independently.

What "verbatim" means

The published text is the submitted text. Sphinnx does not correct spelling, adjust punctuation, change capitalisation, condense, or reorder the author's words. If the author makes a typo, the typo is published. If the author changes their mind mid-sentence, that change appears as written.

The only exception is rendering: line breaks are preserved; HTML special characters are escaped for safe display.

What the markup declares

Every commentary block carries the following signals, visible to readers, search engines, and LLM crawlers:

SignalValueWhere declared
Input methodtyped-directly-no-paste-no-aidata-integrity attribute, meta tag, Schema.org description
Publication ruleverbatim and in fullintegrity-anchor tooltip, Schema.org Answer description, JSON-LD
Author attributionNamed author + date of answerVisible byline, Schema.org Answer.author, JSON-LD
Policy referenceThis page URLmeta name="express-interview-policy", integrity-anchor href

What is and is not covered by this policy

ElementCovered by this policy?
Author's Commentary answersYes — verbatim, typed directly, no AI
Article body textNo — governed by the AI disclosure notice on each article
Questions (AI-generated)No — questions are AI-generated; the author may request one regeneration
Article title and subtitleNo — author-authored but not subject to verbatim rule
Author card bioNo — author-authored, may be edited by author at any time

Moderation and corrections

Commentary answers are not moderated before publication for trusted authors (authors with three or more approved articles). For new authors, the full article including commentary is reviewed within 48 hours of submission.

If an author believes their published answer contains an error of fact, they may submit a correction request. Corrections are published as a new answer appended below the original, clearly labelled as a correction with date. The original answer is never altered or removed.

Contact

Questions about this policy: i137@proton.me

Policy version 1.0 · Published 27 June 2026 · About Sphinnx