Editorial Policy

Topic selection

When you open an OKX tutorial, how do you decide whether to trust it? Here are the four rules we audit ourselves against, written so you can hold us to them.

1. Verifiable against public data — if a claim cannot be reproduced from a public dataset, we do not publish it.
2. Useful to long-term readers — "what is pumping today" stories are not on our list.
3. Focused on OKX in practice — tutorials for other exchanges are out of scope.
4. Beginner-friendly — every piece must let a complete crypto newcomer follow along step by step.

What we never write:

• Price predictions or "potential analysis" of unlaunched projects;
• Airdrop farming routes and interaction guides;
• "10x coin picks" or meme-coin opportunities;
• White-paper-style ghostwritten project intros;
• "Join the group for free signals" invitations;
• Any variant of the "follow the master's trades" model;
• Hands-on guides for other exchanges (focus principle).

Fact-checking

Before anything is published, it goes through two checks:

1. Sources attached — every specific number, date, URL and screenshot is marked, with the original source attached to each one.
2. Checked against official pages — each number is traced back to OKX's own pages or the original data, and each URL is opened to confirm it still resolves.

Anything the check cannot reproduce is rewritten or removed. "Looks roughly correct" is not allowed.

Source labelling

The "Sources" section at the end of every article must list every citation in full, including:

• OKX official docs, help-centre pages and API reference;
• OKX public Proof-of-Reserves reports;
• Public data APIs (CoinGecko, Glassnode, on-chain explorers);
• Academic or institutional research papers;
• Event reporting from established financial press.

Corrections workflow

1. Receive — readers email factual errors to our inbox.
2. Verify — we confirm within 7 days.
3. Correct — factual errors (numbers, dates, outdated screenshots) are fixed in the original article and a revision note is added to the top. Differences of interpretation are not silently rewritten; we append a "Reader response" block at the foot of the page instead.
4. Be transparent — for major errors that affect an article's core conclusion, we also post a notice on the home page on top of the above.

Conflicts of interest

The full financial disclosure lives in section 9 of the Disclaimer. Our process:

• Active disclosure — any tokens, equity or advisory relationship a team member receives, however small, is logged internally.
• Passive avoidance — members holding an asset or tied to a project do not write about that asset or project.
• Because of our OKX affiliate relationship, we deliberately avoid exclusive claims such as "OKX is absolutely better than X."

Writing style

CryptoDesk deliberately keeps a "field manual" tone:

• Second person ("you"); we avoid "we believe / the author thinks / our editors say";
• Short sentences, clear steps, precise numbers;
• One idea per paragraph;
• Lists when they help, never as filler;
• No stock filler phrases — "in today's era…", "all things considered", "needless to say" are removed on sight;
• No fabricated data — a "5/5 rating" must always be backed by a stated methodology.

Reader feedback

Reach us at [email protected]. We reply roughly every two to three days.

We welcome: factual corrections (down to a single decimal point), tutorial requests, feedback on this editorial policy, and applications to become a volunteer reader-proofreader (unpaid; we list contributors in our annual thank-you page).

We do not reply to: paid promotion, signal-group invites, requests to add us on chat apps, or project PR pitches.