Day trading / simulation
Time Machine
Replays the rest of the session 10,000 times plus every past time the setup fired.
Simulates 10,000 sessions from the stock's own recent 5-minute bars and re-runs the same rule on past sessions, then reports target-first, stop-first and still-open odds.
- Replays
- 10,000
- History
- 59 sessions
- Seed
- Repeatable
- Output
- Odds
Live scene
Time Machine
git clone https://github.com/Jakeschincariol/day-trader-skill.gitMITOverview
What Time Machine actually does
The Time Machine is a reality check before risk. It replays the rest of the session 10,000 times, drawing each 5-minute candle from the stock's own bars at that time of day over the previous 59 sessions, with a seed anyone can repeat.
It also re-runs the same rule on those real past sessions and reports how often the target hit first, the stop hit first and the trade was still open at the close. The odds are a simulation, not a prediction.
Capabilities
What it helps your agent ship
Run 10,000 simulated sessions from the stock's own bars.
Replay the same setup on past real sessions.
Report target-first, stop-first and still-open percentages.
Use a seed so the result is repeatable.
Compare the simulation with the historical base rate.
Write the odds for the risk and budget skills to read.
Workflow
The process the skill follows
A repeatable sequence keeps the agent from jumping straight into edits before it understands the job.
- 01
Load the plan
Read the entry, stop and target from trader/plan.json.
- 02
Collect the bars
Pull the stock's recent 5-minute bars by time of day.
- 03
Simulate the session
Draw 10,000 paths from those bars with a fixed seed.
- 04
Replay history
Apply the same rule to every past session where it fired.
- 05
Report honest odds
Show both results and label the simulation as a simulation.
Inside the pack
References and modules
These are the load-on-demand files or command groups that make the skill useful beyond a single prompt.
timemachine.py run --write
10,000 replays plus the history.
--seed 19
Repeatable simulation anyone can rerun.
Same-time-of-day bars
Candles drawn from the stock's own recent sessions.
Historical replay
The rule applied to real past sessions.
Odds report
Target first, stop first and still open at the close.
Reality-check language
No promise and no prediction.
Use cases
Where it earns its place
Pre-trade reality check
See whether a setup is close to a coin flip before risking money.
Setup comparison
Compare two rules on the same stock with the same seed.
Expectation setting
Separate a good story from honest odds.
Prompt starters
Try one of these first
Copy a prompt into your agent. Replace the example URL with your own project or page.
Replay this trade 10,000 times and tell me the honest odds.
Run the time machine with seed 19 and compare it with the historical base rate.
Is this setup actually better than a coin flip?
Keep exploring
Related skills
Want a site that uses this skill properly?
We build Next.js websites with technical SEO, structured data, 3D scenes and motion that stays inside a performance budget.