What this site is
Most material about trading automation stops at the point where the software is chosen. That is where the actual work starts. Someone has to decide which key signs what, move value into wallets without losing track of it, prove a configuration before it costs anything, watch a run that may be silently failing, and take everything back out at the end. This site is about that half of the problem.
The content is organised as runbooks. A runbook states its preconditions, gives numbered steps in the order they must happen, names the step that cannot be reversed, and finishes with a verification you can actually perform. If a page here cannot be followed by someone who is tired and slightly worried, it has failed at its job.
Who writes it
Pages are published under the byline of the Volume Ops Desk. The desk is the editorial function of this site, not an individual, and it deliberately does not invent a personal biography, a photograph or a credential list to make procedures look more authoritative than the reasoning behind them. A procedure is worth following because the mechanism it describes is correct, not because of who signed it.
That also sets a limit the desk holds to: we do not publish trading results, performance claims, case studies or anonymised client stories. There is no way for a reader to verify any of that, and in this niche unverifiable numbers are the main thing being sold. Where a number appears on this site it is either a protocol constant with its formula shown, or arithmetic labelled as illustrative so you can substitute your own inputs.
Who it is for
The reader we write for is the person who operates the software and carries the consequences. That might be someone running an open-source script on a rented server, someone using a hosted console, or someone who inherited a setup from whoever built it and now has to keep it alive. The common thread is responsibility for keys and for value, not the specific tool.
It is not written for people looking for a strategy, a signal or a way to make a chart look busy. Nothing on this site tells you what to trade or when. The procedures assume you have already made those decisions and now need the operational layer underneath them to be boring and repeatable.
How the sections fit together
There are three sections and they map onto three different kinds of consequence.
- Key hygiene covers custody: the threat model for a key a program can sign with, how balances are segmented so one compromise cannot reach everything, and how keys are stored and rotated. These are decisions made once that stay with you for years.
- Runbooks covers procedures performed under time pressure: funding a fleet, the rent and account costs that fleet locks up, the preflight gate before a live run, and a clean shutdown.
- Monitoring covers how long a problem runs before anyone notices: what to alert on, what to log, and the ordered response when a key or an endpoint stops being trustworthy.
What is deliberately out of scope
The desk does not review or rank products. Where a commercial console is referenced, it is referenced as an example of the hosted version of a workflow and it is labelled as a third-party product every time. The desk does not audit those products, does not test their claims, and does not have visibility into how any of them handle keys internally. Reading a provider's own documentation about key handling before funding anything is the reader's job and we say so on the pages where it matters.
We also stay out of legal, tax and financial advice entirely. Whether a particular pattern of automated activity is permitted where you live, and what it means for your tax position, are questions for people qualified to answer them. Procedures on this site are operational, and following one correctly does not make the underlying activity appropriate for your situation.
Corrections
Protocols change, defaults change, and a procedure that was right last year can quietly become wrong. If something here is inaccurate, out of date, or unsafe as written, the desk wants to know, and the fastest route is the contact page. Corrections are made to the page itself rather than buried in a note, and the sourcing rules we hold ourselves to are set out in the editorial policy.