Journal: notes on a trading bot that does not trade yet

The journal is a public notebook for the design of Captay. Each note describes behavior the cabinet would have if it were open. The cabinet is not open, so the notes are not field reports. They do not smuggle in a track record by narrating a hypothetical month as if it had customers.

How to read them

Read a note as a constraint on a future machine. When it says a grid stops at the edge of a range, that is a requirement, not a story about a range that was traded. When it says a halt does not rewind a fill, that is a limit on what the word stop is allowed to mean. If a sentence would only be interesting if a profit followed it, the sentence does not belong here.

The three notes linked below are the whole archive of this release. They overlap the product pages on purpose and then go further: one on the shape of an API key, one on how the three strategies behave when the market is uncooperative, and one on loss limits. They are not chapters of a single essay pasted three times. If a paragraph is identical, the build is supposed to fail.

What will not show up

• No equity curve, including a flat one presented as discipline.
• No ranking of DCA against a grid against rebalance.
• No quote from a user, because there is no user of a live cabinet.
• No exchange badge implying a relationship that was never signed.

Dates on the notes are the publication date of this site, not a backdated research program. Later notes can carry later dates when they exist. Until then, the archive is short on purpose: three titles, one publication date, and no implied sequel.