Who it is forA trader who wants a reality check on a setup before risking money on it.
The problemGut feel says a setup looks great; nobody checks how the same rule did on the last 60 sessions.
The outcomeSimulated odds of hitting the target, the stop or neither, plus how the same rule did on past real sessions, and a position sized to your max loss.
What you need first
- Claude Code (or any Claude chat, without the Python tools)
- Python 3.8 or newer; nothing to pip install
- For SEC tools: your name and email in trader/config.json or SEC_USER_AGENT
- For orders only: a brokerage connector such as Alpaca's MCP server, paper account by default
Steps
- Set the plan before entry:
python3 trader-exit/exit.py plan --stop or-mid --target 1R --write - Replay it:
python3 trader-timemachine/timemachine.py run --write - Size the position to your budget and max loss:
python3 trader-budget/size.py --write - Check it against your limits:
python3 trader-risk/check.pyreturns PASS or BLOCK with the reason
Prompt for your agent
Using Jakeschincariol/day-trader-skill, plan a trade on <ticker> with the stop at the opening range midpoint and a 1R target, run the Time Machine, size it for a <budget> account with a <max loss> max loss, and run the risk check. Report the odds and PASS or BLOCK, and do not send any order.
Caveat. The Time Machine is a simulation built from the last 60 days of bars, not a prediction, and a stop can fill well past its price in a fast market. Not financial advice: paper trade first, and remember you can lose money.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for Jakeschincariol/day-trader-skill and "Replay a trade 10,000 times before you take it"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer