# Before the Fill

> A risk check on one trade before the order goes in. Enter the trade - symbol, long or short, entry,
> stop, an optional target, the multiplier and the account size - and the positions already open. A free
> in-browser engine sizes the trade from the stop distance at your risk per trade, trims it to your
> position, open-heat and group limits, and shows the reward to risk, the break-even win rate, the
> expectancy at your own win rate, the open heat after the trade and a losing-streak table, then gates
> the trade: take, resize, wait or skip. A metered risk read keeps to that gate, reads the stop against
> the thesis, checks the book for the same bet twice and writes a plan for after the fill and the
> checks to make before clicking. Derived from @sickn33/risk-manager.

https://before-the-fill.skillsafe.ai/

## What it does

- Size by risk: (account x risk per trade - round-trip commission) / ((|entry - stop| + slippage) x multiplier), rounded down to the size step.
- Caps: position value as a percent of the account; open heat (the risk from each held position's current price, or entry, to its stop) across the book; open heat within one #group tag. The smallest size wins and the page names the limit that set it.
- Reward to risk after slippage, the break-even win rate after commission, expectancy at your win rate (every trade ends at the target or the stop), 1R/2R/3R price levels, and the drawdown after 3, 5, 8 and 10 losses in a row.
- The gate, in order: negative expectancy or a stop too far for one unit -> skip; no heat or group room -> wait; a planned size over the maximum -> resize; otherwise take.
- Free downloads: the price ladder (SVG, PNG), the heat bar (SVG), the book (CSV), the whole check (Markdown, JSON) and a one-line order ticket.

## The paid task

One task, `check`, on gpt-terra, metered per run. The input is string fields (symbol, side, currency, account, entry, stop, target, multiplier, planned_size, group, atr, win_rate, avg_win_r, limits, open_book, thesis, gate, facts); the reply is one JSON object: headline, verdict (take | resize | wait | skip), size, size_why, stop {read: sound | tight | wide | unclear, note}, target {note}, edge, book, events[], plan[{when, do}], before_you_click[], wrong_if. The page re-checks the reply: verdict against the gate, size against the limits, the stop read against the ATR multiple, and any plan step that widens a stop.

Risk arithmetic and process, not investment advice: the read never says where a price is going.

- API tutorial: https://before-the-fill.skillsafe.ai/api.html
- Source skill: https://skillsafe.ai/skill/@sickn33/risk-manager (sickn33/antigravity-awesome-skills, MIT)