Automation In Chat

You describe the goal. The AI writes the code. You keep the decision.

Claude Code, working session
Read the closed trades and write each one to the Google Sheet.
Reading history, mapping columns, appending rows. One tab per strategy.
Add a live dashboard on top.
Done. Win rate and equity curve update on each write.
# journal_writer.py
def write_trade(t):
  row = t.to_row()
  sheet(t.strategy).append(row)
  dash.refresh()
# reviewed, approved, shipped

The fast build loop

Three moves, on repeat, until the tool is real. You prompt the draft, you verify it against real data, you ship a version you can roll back.

ONE
Prompt to product
intent in words becomes working code
TWO
Verify and ground
run it on real trades and files
THREE
Ship and version
a dated build you can trust and revert

The value is speed with control. The AI removes the typing and the boilerplate. The judgment stays with the person.

From one prompt to a version

One intent in plain words turns into a draft, a check, a shipped build. The old way writes every line by hand. This way, the person keeps the decisions.

build cockpit
PROMPT
1
intent in plain words
DRAFT
240
lines proposed
VERIFY
pass
run against real data
SHIP
v3
versioned, reversible
chat_session.py
# you, in plain words
"Read closed trades, one row each."

# the AI, in seconds
def write_row(trade):
    sheet.append(trade.to_row())

# you review, then approve
if looks_right:
    ship()

Built this way, already running

Real automations, delivered in chat, saving hours every week.

Trade journal

Closed trades to a Google Sheet with live dashboards, one tab per strategy.

Control cockpit

A sheet mirroring strategies, executions, terminals and telemetry.

Desktop monitor

A live view of the machine and every running process, at a glance.

Trading system

A full system, built from scratch with AI assistance, data to order.