Pressure-test the trade before you place it.

Eleven editable skills that turn Claude Code or Codex into a trading research partner. Define the setup, replay it against the bars, and review every paper trade.

One-time purchase. You place every trade yourself.

trader-timemachine

One bar touched your stop and your target. Which filled?

Trade replayreplay.py · bundled example
Three five-minute bars of the bundled example trade A long entry at 100 dollars 10 on the 10:05 bar. That bar ranges from 94 to 111 dollars, crossing both the 95 dollar stop and the 110 dollar target. $110 targettarget $100.10 in $95 stopstop 10:0010:0510:10
Assume the bar hit the
Exit fill
$94.90
Net result
−$52.20
Of planned risk
−1.02R

Long 10 shares with $0.10 slippage and $0.01 fees per share. Five-minute bars can’t show which price came first, so the replay reports both outcomes instead of choosing the flattering one.

Editable skills
11
Local Python tools
2
Orders placed for you
0
One time, USD
$59

The bundle

Eleven skills, in the order a trading day runs.

Work through them in sequence or call the one you need. Each skill is a plain-text folder you can read, edit and version alongside your own rules.

01 · Before the open

Prepare

Know what your data can support and what your setup actually says.

  • trader-dataCheck feed quality, timestamps, adjustments and gaps before trusting a signal.
  • trader-playbookTurn a setup into versioned rules with explicit no-trade conditions.
  • trader-scannerScreen for candidates against criteria you state.
  • trader-newsResearch the catalyst and what was known at the time.

02 · At the trade

Plan

Write down the entry, the exit, the size and the limit before price moves.

  • trader-entrySeparate a matched trigger from an entry you could actually get.
  • trader-exitDefine invalidation, targets and time exits.
  • trader-budgetSize a paper trade inside the budget you set.
  • trader-riskReview exposure, constraints and the conditions that end your day.

03 · After the close

Test and review

Replay the plan, run a paper test, and keep the lesson.

  • trader-timemachineReplay a plan against supplied bars and explore scenarios from your session results. Includes two Python tools.
  • trader-paperFreeze a forward-test protocol before you see the results.
  • trader-journalCompare each decision with its plan and choose one improvement.

Time Machine

Two tools that run on your machine and show their arithmetic.

Python 3.9 or later, standard library only, no network requests. You supply the bars or the session results. They return JSON you can keep with your notes.

A gap through the stopreplay.py
A replayed trade where price gaps through the stop Price drifts down toward a 95 dollar stop, then the next bar opens at 92 dollars 40. The replay fills the stop at 92 dollars 30, a loss of 1.53 times the planned risk. $95 stopstop Filled $92.30
Planned risk
$51.00
Net result
−$78.20
Of planned risk
−1.53R

Replay one trade against the bars

A long or short plan, bar by bar: delayed entry, slippage, fees, gaps through the stop, and bars that leave the outcome ambiguous.

Equity over 20 sessionsscenarios.py
Sixty resampled equity paths over 20 sessions Paths start at 1,000 dollars and spread out. Across all 1,000 paths the 5th percentile ends at 970, the median at 1,030 and the 95th percentile at 1,090. $1,090 $1,030 $970 Session 1Session 20
Median ending equity
$1,030
Median max drawdown
$24
Paths ending below start
17.6%

See the spread your own results imply

Resample your net session results with a seeded block bootstrap. Shown: 60 of the 1,000 paths from the four-session sample input in the documentation.

Neither tool is a strategy backtester. They do not find signals, forecast prices or show that a strategy has an edge; they make the assumptions in your plan visible.

Boundaries

A research partner, not an autopilot.

The skills are built around what an AI agent can responsibly do for a trader, and they are explicit about the rest.

No order placement

Nothing connects to a broker or submits an order. Every trade is yours to place.

No signal feed

Nothing tells you what to buy. The skills work on setups and criteria you define.

No promised edge

Replays and scenarios test assumptions. They cannot predict returns or prove an edge.

Day Trader Skills

The whole workflow, for one price.

$59 one time

  • All 11 editable skill folders
  • Time Machine replay and scenario tools
  • Example input, input formats and methodology notes
  • MIT licensed, so you can adapt every rule
Get all 11 skills

Secure checkout by Whop. No subscription.

Questions

Before you buy.

What exactly am I buying?

Eleven skill folders: instructions and references that guide your AI agent through trading research, planning and review. The Time Machine skill also includes two local Python tools. It is a downloadable toolkit, not a standalone trading app.

Does it trade for me?

No. The skills cover research, trade planning and local paper testing. They do not submit orders or connect to a brokerage account.

Will it tell me which trades will win?

No. Replay and scenarios help you inspect assumptions and explore outcomes. They cannot predict the future or establish that a strategy is profitable.

Is the Time Machine a backtester?

No. One tool replays a single supplied trade plan against OHLC bars, including costs, gaps and ambiguous exit ordering. The other resamples your supplied net session results into equity and drawdown scenarios. Neither searches history for signals or runs a full strategy test.

Do I need a paid data feed?

Not to explore the workflows. You can bring local exports or use data your agent can already reach. Serious testing needs data suited to your strategy and timeframe. Data subscriptions and AI usage are separate costs.

Which markets and styles does it cover?

Intraday setups you define for US stocks and ETFs, long or short. Short planning asks for borrow and cost assumptions. Options, futures and crypto are outside this bundle.

Give the next plan a harder look.

Get all 11 skills $59 one time