What is bulk sentiment tagging and when should I run it?

Retroactive sentiment tagging finds past trades that have a note but no emotion tag and classifies each one. It runs up to 50 at a time to cap the cost, shows the running total in dollars while it works, and can be re-run to continue through the backlog.

It exists because the emotion tags are what make the Psychology analytics work, and almost nobody tags every trade as they go.

What it selects. Trades with a note written and no emotion tag set. A trade with no note is skipped — there is nothing to classify. It uses the same prompt as the per-trade Auto-tag button, so the results are consistent with anything you tagged by hand that way.

What it costs. Batches of up to 50, deliberately, so a run has a bounded price. The card shows the candidate count before you start, a live counter while it runs — processed, tagged, dollars to four decimal places — and a summary at the end with the total cost and token count. On a cheap model this is cents.

It gets cheaper if economy mode is on. Bulk classification is the one case where the app downgrades even a paid user's model choice, because a cheap model classifies sentiment as well as an expensive one.

When to run it: once, after you have been journalling long enough to have a backlog worth analysing, and then again whenever you have let tagging slide for a few weeks. Re-run it until the candidate count reads zero.

Note that not every trade gets tagged. The summary separates tagged from skipped (no emotion signal) — a note reading "filled at 5" contains nothing to classify, and the model correctly declines rather than inventing a feeling.