How to Use Day Trader Skill: A Local Paper-Trading Tutorial

This tutorial uses the original GitHub repository's local Python tools. It follows the order of the project's October 19, 2026 Tesla example: scan, news, entry, exit, simulation, share size, risk check and paper ticket. The quoted prices and results belong to that dated example. A new run uses the data available when you run it.

Before you start

The commands below require a local copy of the original GitHub repository and Python 3.8 or newer. Open a terminal and run these commands in order:

git clone https://github.com/Jakeschincariol/day-trader-skill.git
cd day-trader-skill

If the repository is already on your computer, open that folder instead of cloning it again. Run the remaining commands from the repository root. Before sizing a proposed order, create the local settings file from the source project's Budget template:

mkdir -p trader
cp -n trader-budget/config.example.json trader/config.json

Open trader/config.json and set your own budget, maximum loss per trade and other limits. The template starts in paper mode. cp -n leaves an existing settings file untouched. Keep the trader/ files together so each tool can read the previous step's output. For SEC requests, replace the template's sec_user_agent placeholder with your name and email, or set SEC_USER_AGENT in your own environment. Keep brokerage credentials in your own environment; this walkthrough ends before order submission.

Installed only the Claude Code plugin? Open the Skill guides and use their Use with Claude requests. Plugin installation does not create the cloned working folder required by the terminal commands on this page.

1. Scan for a setup

python3 trader-scanner/scan.py orb --at 10:15

In the dated source example, the scanner found 42 opening-range matches among 231 usable charts. First read the as of time, attempted and usable chart counts, and any PARTIAL marker. A hit means a completed five-minute close met a rule. It is not a selection to buy. For a controlled study, follow the source's saved historical input rather than treating a fresh live run as the same experiment. Scanner guide.

2. Inspect news and filings

python3 trader-news/news.py --tickers TSLA --hours 12

Check the source and New York time of each item. The historical README describes a TSLA headline and an upcoming earnings context; it does not establish today's news. If SEC items are included, provide the required SEC identification in your own local environment. A headline classification is context, not proof of a price direction. News guide.

3. Check the entry condition

python3 trader-entry/entry.py check TSLA --write

The project's example was waiting at 10:12 because the relevant bar was still forming, then triggered at 10:15 after a completed bar closed above the opening range. Read the state and timestamp together. --write saves the local entry file for the next step. If a current run has a different price or state, do not replace it with the historical example's values. Entry guide.

4. Define the exit before sizing

python3 trader-exit/exit.py plan --stop or-mid --target 1R --write

For the historical source input, entry 364.18 and midpoint stop 358.32 imply $5.86 planned risk per share; the 1R target is 370.04. --write stores trader/plan.json. Read the entry, stop method, target method, and flat-by time from your own output. The stop is a plan; a live fill can slip beyond it. Exit guide.

5. Compare simulation and history

python3 trader-timemachine/timemachine.py run --write

The source's dated example reports 29.4% simulated target first, 25.5% stop first, and 45.1% neither by the close. Separately, its past-rule check found 33 occurrences across 59 sessions: 13 target first, 15 stop first, 5 neither. State the sample and method beside every percentage. These numbers are neither a forecast nor a profit estimate. Time Machine guide.

6. Calculate the proposed whole-share size

python3 trader-budget/size.py --write

The source example uses a $5,000 account and $50 maximum planned loss per trade. With $5.86 planned risk per share, the whole-share limit is eight; at the planned stop, that is $46.88. Your own cash, position cap, and config may make the permitted quantity lower or zero. --write saves a local proposal, not a broker order. Budget guide.

7. Check the configured rules

python3 trader-risk/check.py

Read each PASS or BLOCK reason and the data used. A pass only says the proposed order met those configured checks at that moment; it does not mean the price is fresh enough to trade or that the setup is worthwhile. Issue #32 reports one case of a market-order check passing after a fresh price had crossed a planned stop. Check the current issue status and broker quote before relying on the path. Risk guide.

8. Review a paper ticket without submitting it

python3 trader-broker/ticket.py show

Inspect ticker, side, quantity, planned entry, stop, target, paper/live mode, and readiness. show displays a local ticket; this tutorial does not provide an approval or submission command. The source's actual broker path is separate and requires an exact-order risk pass, explicit typed approval, and a Claude tool-permission confirmation in a correctly configured environment. This site cannot place an order. Broker guide.

Common mistakes to avoid

Mistake	Better reading
Treating a scan hit as a pick	It is a rule match to investigate.
Treating the dated TSLA numbers as today's prices	Keep the 2026-10-19 example date beside them.
Calling a forming bar a trigger	Wait for the completed-bar state.
Combining simulated and historical percentages	Show their methods and denominators separately.
Assuming a stop or PASS caps real loss	Check slippage, gaps, current broker quote and permissions.
Treating `ticket.py show` as an order	It is a read-only ticket display.

Common questions

Can I follow the exact historical figures with a current live run? No. Current data and configuration will differ; use the source's saved fixtures for a reproducible historical study.
Do I need to install each skill separately? No, a complete pack installation already includes all 11.
Does this tutorial execute a trade? No. It ends at a displayed paper ticket and contains no website trading action.

Sources

Original README and historical walkthrough, checked 2026-11-05 against main commit 3d0f4ef.
The linked original SKILL.md on each step's Wiki page describes tool input, output and limits.
Issue #32 is a user report, open at the 2026-11-05 check.