Casework: How This Desk Works

What the desk covers

Trace Rack writes about following value on Solana: how to read a transaction from its own record, how to walk back to a first deposit, how to group addresses with a stated confidence, how to follow a token supply hop by hop, and how to recognise the structures where all of that stops working. Eight guides cover the ground, split between the procedural half and the half that keeps the procedural half honest.

The audience is researchers, analysts and traders who want the method rather than a subscription to somebody else's conclusions. Everything published here can be run against any address by anyone with a public endpoint and patience, which is the standard the desk holds itself to: if a claim cannot be reproduced from public data by a reader, it does not belong on the site.

Who writes it

Pages are published under The Trace Rack Desk, which is the editorial identity of the site rather than a person. There are no author biographies here, no photographs and no invented credentials, because none of those would tell a reader anything useful about whether a procedure works. What tells them that is the procedure itself: the inputs are named, the steps are listed, and the false-positive mode is stated on every method.

Where the material comes from

Three sources, and nothing else. Protocol behaviour comes from public documentation for Solana and the SPL token program, and is described in terms that a reader can verify against the same documents. Method comes from procedures the desk can run itself against public data, written up as steps rather than as results. Anything attributed to a third party is quoted with its source named, and treated as that party's claim rather than as an established fact.

Numbers appear only when they are protocol constants, such as the lamport denomination or the rent-exempt minimum for a token account, or when they are clearly labelled illustrative arithmetic built from constants and invented round figures. There are no accuracy rates, cluster counts, success percentages, review scores or any other statistic on this site, because the desk has not measured any such thing and does not publish numbers it did not observe.

The ethical boundary

This is a site about reading a ledger, not about identifying people, and the distinction is enforced rather than mentioned. Addresses are public keys. The record shows that a key signed a transaction; it never shows who held that key, whether access was shared, or whether the key changed hands between two transactions. Every method here stops at that boundary and says so in the page.

Concretely, the desk does not publish techniques whose realistic use is de-anonymisation: correlating on-chain timing with social media activity, building identity indexes from name services and profile links, joining chain data to breached datasets, or any form of pretexting against a service. It does not publish accusations against named wallets, projects or individuals, does not publish clusters of real addresses, and does not use labels such as scammer, insider or launderer, which assert conclusions that public data cannot support.

Examples use placeholder labels written as letters rather than base58 strings, so no page here points at a real account. The reason is not squeamishness. A worked example built on a real address turns a teaching illustration into a permanent public claim about somebody, and the claim would keep circulating long after the example stopped being accurate.

What the desk refuses

• Tracing work for clients, of any kind, paid or unpaid.
• Investigations into a specific wallet, token, project or person.
• Requests to identify who is behind an address.
• Publishing address lists, clusters or watchlists.
• Verdicts on named products, including the tools the site references.
• Any content whose main function is to direct pressure at somebody.

How findings are checked

Every procedure published here is written so that it can be broken. That means the page names the ordinary situation that produces its signal by accident, states the strongest wording the evidence supports, and gives the reader the check that would show the method failing. A method with no stated failure mode has not been examined properly, and the desk treats an unstated failure mode as an error rather than as a style choice.

Where a page contains arithmetic, the inputs are stated so a reader can redo it. Where a page contains a table of signals, the weights are ordinal judgements and are labelled as judgements rather than measurements. Where a page describes a limit, the limit is a property of the system rather than of current tooling, and the page says which of those two it is.

Corrections

Errors get fixed visibly. When a page changes in a way that alters what it claims, the change is made in the text rather than quietly patched, and the desk would rather carry a note that a previous version was wrong than present a corrected page as though it had always said the right thing. Corrections can be sent through the contact page, and the most useful ones name the page, quote the sentence and say what the record actually shows.

Links to tools

Some pages link to automation products where the paragraph is genuinely about automated trading behaviour, because documented tool behaviour is useful reference material for anybody trying to read a pattern. Those links are ordinary editorial links to public pages. The desk does not review those products, does not rank them, does not publish ratings, and nothing on this site should be read as an assessment of whether any particular tool is suitable for anything.

What this site is not

It is not a service, not a dashboard and not a source of investment or legal advice. It is not a place to report suspected fraud, and it cannot help recover funds. It has no view on whether any specific address, token or project is legitimate, and it will not develop one, because forming that view from public data would require exactly the inferential leaps the whole site argues against.