An existing recorded run

Review before saving.

This viewer presents an already completed offline scenario. Switching steps does not call a model, save new trades, or send messages.

Simulation

The screenshot and all records are synthetic. Extracted fields and the assistant reply are mocked test responses, not Claude output.

Synthetic trade screenshot: BTCUSDT long, entry 100, exit 110, quantity 2, stop 95 — The input from the existing demo. The image explicitly identifies the trade as synthetic.

Mocked extraction response

Not every field is correct.

Instrument: BTCUSDT
Direction: —
Entry: 100
Exit: 110
Quantity: 1
Stop: 95
Date: 2026-01-15 10:00

The mocked response misses direction and gets quantity wrong. There is no saved journal entry yet.

Saved records: 0

The app does not pick long for you.

Confirm direction:
side=long or side=short.

The trade was not saved.

English rendering of the application check from the existing run. The draft remains editable.

User correction

Review. Correct. Confirm.

side=long; quantity=2; stop=95

Instrument: BTCUSDT
Direction: long
Entry: 100
Exit: 110
Quantity: 2
Stop: 95

The next user action in the recorded scenario is to confirm the corrected fields.

This button only displays the previously recorded confirmation result. It does not create a new entry or send data.

Saved in the test journal

A confirmed record, not a guess.

Recorded result: BTCUSDT long, PnL +20, R +2. This is a synthetic trade, not user income.

First, one entry from the screenshot. After confirmation, six further synthetic entries were added separately to demonstrate reflection. Seven entries in total, not seven trades extracted from one image.

1 screenshot entry + 6 additional demonstration entries
Instrument	Side	PnL	R
BTCUSDT	long	+20	+2.0
DEMO1USDT	short	+2	+0.2
DEMO2USDT	long	0	0
DEMO3USDT	short	−2	−0.2
DEMO4USDT	long	+4	+0.4
DEMO5USDT	short	−6	−0.6
DEMO6USDT	long	+8	+0.8