An AI-powered journal that writes itself.
Shifu already holds the alert, the order and the fill. No spreadsheet, no export. Ask Claude or ChatGPT anything about it.
one trade, end to end
journal entry · 09:41:22 CT
illustration
Illustrative journal entry. Strategy MNQ Momentum, buy 2 MNQ, filled at 20,184.25, speed 214 milliseconds, slip plus 0.25 points, plus 1.40 R, closed by the take-profit, plus $112.00. The figures are invented to name the fields.
Written, start to finish
Four stages, no keystrokes.
- alert: The alert arrives. TradingView fires; Shifu takes delivery. Which strategy asked, which side, which contract, how many — and the price the alert carried, the yardstick Slip is measured against.
- order: The order goes out. Shifu forwards that exact order to your broker and decides nothing. The journal clocks what the broker did with it: how long the fill took to confirm once the order was handed over.
- fill: The fill comes back. A real price, from your account. Its distance from the price the alert carried is Slip — the number a statement cannot give you, because only the sender knows what was asked for.
- journal entry: The entry is already written. Realized against the stop recorded on the entry, and which leg closed it. Nothing typed, nothing imported, nothing reconciled — the write-up is what the routing left behind.
Illustrative entry — the figures are invented to name the fields. Yours are computed from your own fills.
Not a screenshot. The journal itself.
Nobody traded this month. The figures are invented so you can read the thing before you own it.
March 2026
Sun
Mon
Tue
Wed
Thu
Fri
Sat
Week
days follow the CME trade date · 5pm CT roll
Record · BE band ±$25
37W 26L 3BE
13W 6L 1BE days · win rate 59% (BE excluded) · decided on dollars
By weekday · $
A month that never happened — the days, the fills and the totals are all invented. On your account the same grid fills itself from the orders Shifu routed, on the CME trade date, with its 5pm CT roll, so it reconciles against a broker statement instead of arguing with it.
Mar 17
+$403.00
3W 1L · 8 fills
- LongMNQ ×2entry
09:41:22 · broker 74 ms · +0.25 pt
Apex 50K EvalMNQ Momentum - Long exitMNQ ×2+$112.00
09:56:10 · TP · paired entry 20,184.25 → 20,212.25 · held 14m 48s · +1.40R
Apex 50K EvalMNQ Momentum - LongMNQ ×2entry
10:18:04 · broker 81 ms · +0.50 pt
Apex 50K EvalMNQ Momentum - Long exitMNQ ×2-$136.00
10:43:42 · SL · paired entry 20,196.25 → 20,162.25 · held 25m 38s · -0.98R
Apex 50K EvalMNQ Momentum - LongMES ×3entry
11:22:07
Paper · practice (practice)MES Fade - Long exitMES ×3+$165.00
12:04:52 · signal · paired entry 5,712.25 → 5,723.25 · held 42m 45s · +1.63R
Paper · practice (practice)MES Fade - ShortMNQ ×2entry
13:08:30 · broker 77 ms · +0.25 pt
Apex 50K EvalMNQ Momentum - Short exitMNQ ×2+$262.00
14:26:11 · TP · paired entry 20,241.75 → 20,176.25 · held 1h 17m · +1.88R
Apex 50K EvalMNQ Momentum
| Time (CT) | Account | Strategy | Trade | Fill | Broker response | Slip | R | P&L · $ |
|---|---|---|---|---|---|---|---|---|
| 09:41:22 | Apex 50K Eval | MNQ Momentum | LongMNQ ×2 | 20,184.25 | 74 ms | +0.25 pt | — | entry |
| 09:56:10 | Apex 50K Eval | MNQ Momentum | Long exitMNQ ×2 | 20,212.25 | — | — | +1.40R | TP+$112.00 |
| 10:18:04 | Apex 50K Eval | MNQ Momentum | LongMNQ ×2 | 20,196.25 | 81 ms | +0.50 pt | — | entry |
| 10:43:42 | Apex 50K Eval | MNQ Momentum | Long exitMNQ ×2 | 20,162.25 | — | — | -0.98R | SL-$136.00 |
| 11:22:07 | Paper · practicePractice | MES Fade | LongMES ×3 | 5,712.25 | — | sim | — | entry |
| 12:04:52 | Paper · practicePractice | MES Fade | Long exitMES ×3 | 5,723.25 | — | sim | +1.63R | signal+$165.00 |
| 13:08:30 | Apex 50K Eval | MNQ Momentum | ShortMNQ ×2 | 20,241.75 | 77 ms | +0.25 pt | — | entry |
| 14:26:11 | Apex 50K Eval | MNQ Momentum | Short exitMNQ ×2 | 20,176.25 | — | — | +1.88R | TP+$262.00 |
Times in CT; the day itself follows the CME trade date. Broker response = from the order leaving Shifu to your broker's confirmation. Slip = fill vs the alert's signal price. Both exist only for orders Shifu routed, so a hand-placed fill reads as a dash. R = realized P&L vs the stop distance recorded on the entry.
Eight invented fills from a day that never traded. The bracket legs the broker filled on its own show a dash under Broker response and Slip: no alert, so no timestamp and no signal price to measure against. The two Practice rows have no broker, so they show no broker response and no slippage. That is missing data shown as missing, and the journal never guesses a number it does not have.
Four numbers a spreadsheet can't give you.
anatomy of one trade
pick a column
journal entry · 09:41:22 CT
illustration
Illustrative journal entry. Strategy MNQ Momentum, buy 2 MNQ, filled at 20,184.25, speed 214 milliseconds, slip plus 0.25 points, plus 1.40 R, closed by the take-profit, plus $112.00. The figures are invented to name the fields.
Broker response · ms
How long your broker takes to confirm each order Shifu sends it: from the order leaving Shifu to the confirmation.
- Broker response (ms): How long your broker takes to confirm each order Shifu sends it: from the order leaving Shifu to the confirmation.
- Slip (pts): Where the fill landed against the price your alert carried. Costs and gifts both shown.
- R (multiple): Realized P&L against the stop distance recorded on the entry — not a stop you typed in later.
- Exit (TP · SL · other): Which leg closed it: the resting take-profit, the resting stop, or a market order.
The same invented trade the hero writes, and the first row of the tape above — one trade, told three ways. Yours are computed from your own fills.
Routing quality
broker response · live orders Shifu routed
Broker response
77 ms
p95 142 ms
Slippage
-$84.50
54 fills vs signal price
Exit mix
TP 34 · SL 22 · other 10
Avg R · win / loss
+1.28R/
losses holding to the recorded stop
Hold · win / loss
12m/
entry fill → exit fill
Expectancy per trade
+$18.83
avg over 66 closing trades
Broker response and slippage exist only for orders Shifu routed from an alert. A trade placed outside Shifu, and a stop or target leg the broker filled, has no alert to time or price against, so it is left out here, not counted as a slow fill. Set your journal to Manual in settings to hide this block.
Journal equity curve · March 2026 · $
best day +$403.00 · worst day -$310.00
Cumulative from the fills in this journal — not your broker's account equity. Anything Shifu cannot see moves the balance without moving this line.
Best day
+$403.00
Mar 17, 2026
Longest green streak
4 days
in a row · this month
Largest win
+$268.00
single trade · this month
Top instrument
MNQ
+$980.00
Invented figures again — but the shape of this panel is the point: broker response and slippage exist because Shifu sent the order, not because anyone typed them in. An importer has nothing to put in these boxes.
Whether it actually paid.
Month breakdown · March 2026 · $ · net
+$1,243.00gross +$1,429.40 before $186.40 fees▲ $757.00 vs Feb
Winners · profit side
$3,980.00Losers · loss side
-$2,737.00Edge quality · does the system pay
+$18.83/tradeAn invented month. What is not invented is the method: every figure here is measured from the fills, and there is no score, no grade and no rating anywhere in the journal — a number Shifu made up about your trading would be worth nothing to you.
Which instrument, which strategy, which account.
By instrument · $
By strategy · $
By account · $
By strategy: each trade is filed under the bot whose alert placed it, and a target, stop or trail leg under the trade it closed. A trade placed outside Shifu, by hand or by another tool, is filed under Outside Shifu and never counts toward a bot. A close by your broker's risk rule (for example a daily loss limit) is filed under the bot whose trade it closed, marked Broker.
When & how you trade · session timing + trade-size shape
By session · CME clock
Trade P&L distribution · $
Invented buckets. What is real is the attribution: the strategy column comes from the alert that routed the order, so it is which of your bots did it — not a tag you remembered to add.
The record keeps itself too.
Every month lands in the record the day it closes.
There is nothing to tally and nothing to carry forward — the red ones land too, which is why the record is worth having.
Season · best-of-3 months · scored in $ net
Winning seasonSeason 1 (Q1) · 2026
2–1
+$987.00
Winning season — closed 2–1.
Jan
Loss-742-$742.00
Feb
Win+486+$486.00
Mar
Win+1.2k+$1,243.00
Career
2 / 3 winning seasons
months 5W 4L · best: S1 ’26 +$987.00
An invented three months, one of them red — the figures here are made up, not a trader's results. The panel above is the real one from your dashboard.
Set once. It stays set.
Everything above is one trader's settings. Traders don't agree on what a number should be — ticks or dollars, Sunday or Monday, one account or all of them. Pick yours and the journal opens that way every time.
Each day also takes your own write-up and tags, so the calendar carries the why and not just the number. Kept in your browser today — cross-device sync is on the list, and we'd rather say so than imply it already works.
- P&L in dollars, ticks, points or R
- Your calendar week start and display timezone
- Scope to any set of accounts and strategies
- Blur dollars, balances and fills with one screenshot-safe toggle
- Switch panels off — the journal is the ones you keep
- Every view is a link, so a shared view opens the same
It journals what Shifu routed.
No CSV import, no statement upload. A trade Shifu never touched is a trade the journal never saw.
What this journal can see
per connected account · verified against the broker layer
Practice
Complete- Paper · practice412 closed
Every practice fill is simulated inside Shifu, so the journal is the whole record.
Differs from a broker: Practice fills a market order right away at your alert's price. A limit or stop order waits until a later alert's price reaches it. Practice fills with no commission, and a stop or target settles at the next alert, so its results can differ a lot from a broker account on the same alerts.
Tradovate
Every fill on the account- Apex 50K Eval268 closed
Every fill on this account is journaled, the orders Shifu routed and the trades you place yourself in any contract. Those read Outside Shifu and never count toward a bot. A close by your broker's risk rule reads Broker and counts toward the bot whose trade it closed. A trade placed outside Shifu appears within a few minutes. Balance comes from Tradovate.
Not captured: Tradovate lists a session’s orders only until the session ends, so trades from before this account was connected, or from a session it spent needing to sign in again, are not in the journal.
Rithmic
Entries and fills only- Funded · Rithmic96 closed
Orders Shifu routes to Rithmic are journaled with broker fill timing. Exits and realized P&L are not booked yet, so wins, losses and P&L stay empty for these accounts.
Not captured: Orders placed by hand in your Rithmic platform are not journaled, and account balance is not refreshed from Rithmic.
ProjectX
Not reading yet- ProjectX eval0 closed
The ProjectX integration is not live, so nothing is read from this account.
Not captured: No orders, fills, or balances are journaled for ProjectX accounts.
The journal equity curve is built from the fills in this journal, not from your broker's account equity. If you trade outside what Shifu can see, your balance can move while this curve does not — the balance tile is the broker's number, the curve is the journal's.
Four sample accounts, one per route, so every level shows at once. The wording is the product's own — what changes on your dashboard is the account names and the counts.
Which route carries your account? See the broker page.
Included in the plan
The journal isn't an upsell. It's what routing leaves behind.
Start on paper and the calendar fills in from your first simulated fill — no broker login, nothing at stake, the same journal you'll keep when it's real.