From a sentence to a tested result.

Vaquero implements and tests. It does not originate. That split holds at every step below and it is the reason the rest of the page can be specific: once the strategy is yours, the only interesting questions are whether it was implemented faithfully and whether the test was honest.

So nothing here proposes a strategy, ranks one against another, or tells you what to hold. Vaquero publishes no signals, no picks and no model portfolios, and it is not a broker. Nothing you build here places an order or holds your money.

Plain words, in whatever order they come out. There is no syntax to learn and no menu of blocks to assemble, because what reads it is a language model and not a form.

A description is complete once it answers four things. Everything else can be left for the agent to ask about.

The AI agent reads your description and writes it out as a structured, executable specification: the universe, the conditions that open and close a position, the rebalance cadence, and the execution assumptions the run will be charged under. That document is the thing the engine actually runs, and you can read it before it runs.

Where a missing choice would move the result, the agent asks instead of filling it in. A description that does not say what happens on a tie, or over how many sessions an average is taken, has two possible readings and they do not produce the same curve. Guessing at that point is how a backtest ends up answering a question nobody asked.

It also refuses. When a request cannot be expressed exactly as written, the run stops and names the reason rather than running something close to it: a cadence finer than the data's own grid, a window shorter than the signal's own lookback, a universe too wide to price. A refusal is one thing to change and try again. A quiet substitution is a curve that answers a different question in the same font.

The specification becomes a sequence of decision instants. At each one the strategy is handed the bars that had closed and the values that were public on that date, and nothing that was revised afterwards. It cannot see the price it is about to be filled at.

Orders are filled in the session the decision was made in, and that session charges for them: the round-trip fee, the slippage, the funding a perpetual position owes at each settlement, and the push the order itself puts into the book, which is priced as a cost of its own rather than assumed away. Crypto and US equities run through the same engine on the same terms.

The engine page sets out what a decision can see and how a fill is priced. The methodology page sets out the guards that keep the first of those true.

The equity curve against the benchmark you chose, where the money went, and what the book held on its last day. For a run whose instruments you named, the fills come back with their price and time, and pressing a point on the curve brings back the holdings at that decision with the timestamp of the decision itself.

Beside it the run prints what the answer cost to get. The execution assumptions, and the sample length, skewness and kurtosis measured from the run's own returns. A Sharpe out of a short, lopsided record and a Sharpe out of a long well-behaved one are not the same claim, and the result carries what tells them apart. Survivorship comes off the grade as a deduction and the size of that deduction is printed too.

Six settings, and they belong to the run rather than to the account. They are stored with the result, so a run can be re-read later on the terms it was made under instead of on today's.

The same line is worth stating at the end as at the start, because the workflow above is easy to mistake for a service that hands you positions.