The Day Trader skill is a free pack of 11 Claude skills for day trading. It scans charts, tracks news, times entries and exits, and replays a trade 10,000 times before you risk a rupee or a dollar. It costs nothing, needs no API key, and installs in about 5 minutes. This guide shows every step, on Windows or Mac.
It is a Claude Code skill pack by Jake Schincariol, and it went viral in October 2026 as the “Time Machine” skill. I read the repo before writing this. Below is what it really does, how to set it up, and what it can’t do.
Key takeaways
- It’s 11 skills in one free, MIT-licensed GitHub repo: scanner, news, entry, exit, Time Machine, risk, budget, broker, alerts, analyst and a Congress tracker.
- Setup is one paste into Claude, or three commands. Python 3.8+ is the only requirement. Nothing to pip install.
- Nothing can place an order without your typed “yes”, and paper trading is the default.
- The Time Machine is a simulation, not a prediction. In the author’s own Tesla example it came out as “close to a coin flip”.
Not financial advice. This post explains how to install and use software. It doesn’t tell you what to trade. Most people who day trade lose money, and you can lose money.
- What is the Day Trader skill?
- What do you need before you start?
- How do you install it?
- How do you set your limits?
- What does each of the 11 skills do?
- How does the Time Machine work?
- How do you run one trade, start to finish?
- How does it keep your money safe?
- What can it not do?
- FAQ
What is the Day Trader skill?
It’s an open-source pack of 11 Claude skills for day trading, published on GitHub under the MIT licence. Each skill is a small folder with instructions for Claude and a plain Python tool. The tools read free public data: Yahoo Finance charts, Nasdaq Trader symbol lists, SEC EDGAR and the US House Clerk.
Each skill has one job, and each one reads what the last one saved. They write to a trader/ folder in your project. The source is at github.com/Jakeschincariol/day-trader-skill.
What do you need before you start?
You need Claude (Claude Code is best) and Python 3.8 or newer. That’s it. The README says there’s nothing to pip install, because the tools use only Python’s standard library. There’s no signup and no API key for the data.
Two optional extras. The SEC tools (news filings and the earnings analyst) ask you to identify yourself with a name and email. And if you ever want to send orders, you need a brokerage connector. The pack points to Alpaca’s MCP server, with its paper account as the default.
Don’t have Claude Code? You can still paste any single SKILL.md at the top of a chat and it runs as a mode. You lose the Python tools, so the method works but the live data doesn’t.
How do you install the Day Trader skill?
There are three ways. Pick one. The first is the easiest and works for most people.
Option 1: paste it into Claude (easiest)
Open Claude and paste these two lines exactly:
https://github.com/Jakeschincariol/day-trader-skill install skill
Option 2: install as a Claude Code plugin
Run these two commands in Claude Code:
/plugin marketplace add Jakeschincariol/day-trader-skill /plugin install day-trader-skill@day-trader-skill
Installed as a plugin, the skills get a prefix, like /day-trader-skill:trader-scanner. If you want the short /trader-scanner form, use Option 3.
Option 3: copy the folders yourself
On Mac or Linux:
git clone https://github.com/Jakeschincariol/day-trader-skill.git cp -r day-trader-skill/trader-* ~/.claude/skills/
On Windows (PowerShell), the same idea:
git clone https://github.com/Jakeschincariol/day-trader-skill.git Copy-Item -Recurse day-trader-skill\trader-* "$env:USERPROFILE\.claude\skills\"
Want it for one project only? Copy the same trader-* folders into that project’s .claude/skills/ folder instead.
Check that it worked
Restart Claude Code, type /trader- and you should see the skills in the list. If you installed as a plugin, type /day-trader-skill: instead. On Windows, if a tool says python3 isn’t found, ask Claude to run it with python instead. That’s how Python is usually named on Windows.
How do you set your limits and your name?
Set your limits before your first trade. The pack ships an example config, config.example.json, in the trader-budget folder. These are its defaults:
{
"budget": 5000,
"max_loss_per_trade_usd": 50,
"max_daily_loss_usd": 150,
"max_trades_per_day": 3,
"min_reward_risk": 1.0,
"require_stop": true,
"no_entries_before": "09:45",
"no_entries_after": "15:30",
"flat_by": "15:55",
"broker": "alpaca",
"paper": true,
"allow_options": false,
"allow_margin": false,
"allow_short": false,
"sec_user_agent": "Your Name [email protected]"
}Save your version as trader/config.json in your project. Change the budget and loss limits to numbers you can afford to lose. Put your own name and email in sec_user_agent, because the SEC asks automated tools to identify themselves. Leave "paper": true on.
The Risk Manager checks every trade against these numbers, so they work as your rulebook. Set them while you’re calm, not in the middle of a trade.
What does each of the 11 skills do?
| Command | Job | What it does |
|---|---|---|
/trader-scanner |
Chart Scanner | Scans thousands of charts for setups like opening-range breakouts, flags and oversold bounces |
/trader-news |
News Monitor | Tracks breaking news on stocks you watch and flags what moves them |
/trader-entry |
Entry Timer | Watches a setup bar by bar and tells you when it triggers |
/trader-exit |
Exit Planner | Sets your stop and target before you enter |
/trader-timemachine |
Time Machine | Replays the trade 10,000 times and over every past session |
/trader-risk |
Risk Manager | Checks each trade against your limits and blocks anything over |
/trader-budget |
Budget Manager | Sizes the position to your budget and max loss |
/trader-broker |
Order Executor | Sends approved orders to your broker, paper by default, only after your yes |
/trader-alerts |
Price Alerts | Rings when a price, stop or the opening bell hits |
/trader-analyst |
Earnings Analyst | Reads earnings reports and sums up what changed |
/trader-pelosi |
Pelosi Tracker | Reads Congress’s public stock-trade disclosures when they’re filed |
You don’t need all eleven on day one. Start with the scanner, the exit planner and the Time Machine, then add the rest.
How does the Time Machine skill work?
The Time Machine replays the rest of a trading session 10,000 times before you enter. Each replay is built from the stock’s own 5-minute candles over the previous 60 days, matched by time of day. It also re-runs your exact rule on each of those real past sessions. Then it prints the odds of hitting your target first, your stop first, or neither.
The repo’s worked example uses Tesla on Monday 19 October 2026, with seed 19 so anyone can repeat it. The result: target first 29.4%, stop first 25.5%, still open at the close 45.1%. On the 59 real past sessions, the rule fired 33 times: 13 hit the target, 15 hit the stop, 5 neither. The tool’s verdict was “close to a coin flip” (Day Trader skill README, October 2026).
That’s the point of the skill. It’s not a crystal ball that finds winners. It’s a reality check that tells you plainly when a setup has no edge. A tool that says “coin flip” and saves you from a bad trade is doing its job.
Two limits to know. It uses only the last 60 days, so it can’t see an earnings report, a trading halt or a surprise headline that those days never had. And it checks 5-minute bars, not every tick, so real fills can be worse than the replay.
How do you run one trade, start to finish?
The skills run in a chain, and each reads the file the last one wrote. The README gives this order for one trade: scanner, news, entry, exit, Time Machine, budget, risk, broker. Alerts then watch the trade while it’s open.
/trader-scannerfinds stocks that broke above their opening range./trader-newschecks the day’s headlines for the stock. Skip anything with an earnings report due./trader-entrywatches bar by bar and says “waiting” or “triggered”./trader-exitsets the stop and target before you enter./trader-timemachineshows the odds. If it says coin flip, you can stop here./trader-budgetsizes the position. With a $5,000 account and a $50 max loss, the example risks $5.86 a share, so it sizes 8 shares./trader-riskreturns PASS or BLOCK with a reason./trader-brokershows you the ticket in plain words. Nothing is sent until you type yes./trader-alertsrings on your stop, your target, and at 3:55 PM to flatten what’s left.
The tools also run straight from the command line if you prefer, for example python3 trader-scanner/scan.py orb --at 10:15 for opening-range breakouts so far.
How does it keep your money safe?
By design, no trade goes through until you say yes. Only /trader-broker can lead to an order, and only after /trader-risk passes that exact order and you type yes for it in the chat. Paper trading is the default, and broker keys belong in environment variables, never in the chat.
The README adds a second lock. Set your brokerage tool to “Needs approval” in Claude (Customize, Connectors, tool permissions), never “Always allow”. Then Claude’s own Allow once / Deny card asks you one more time before any order leaves.
My advice: stay on paper trading for at least a few weeks. Learn what the tools tell you. Only think about live money once you trust your own process, and only with money you can afford to lose.
What can the Day Trader skill not do?
It isn’t a Bloomberg Terminal, and it doesn’t predict the market. It reads free, unofficial data. Yahoo Finance’s chart endpoint can lag, change or rate-limit you, so check anything important against your broker.
It also won’t make day trading profitable. A simulation built on the last 60 days can’t see the future. Day-trading gains are taxed as short-term gains, and frequent trading can trigger wash-sale rules. And the US pattern day trader rule changed: FINRA removed the $25,000 minimum on 4 June 2026, but brokers may phase it in until 20 October 2027, so yours may still apply the old rule (Day Trader skill README). If you’re in India, check the rules for your broker and market first. This tool reads US data.
Want skills like this built for your business?
Skill packs like this one are a good model for how AI should work: one clear job per skill, with a human saying yes. That’s how I build marketing workflows too. I’m an AI-native digital marketer, and I set up the SEO content pipelines, ad reporting and Instagram automations I recommend. See my AI-powered digital marketing services or browse the projects I’ve built.
FAQ
Is the Day Trader skill really free?
Yes. It’s MIT-licensed, with no signup and no API key. The tools read free public data. Claude itself may need a paid plan to use Claude Code, depending on your account.
Do I need to know Python or coding to set it up?
No. Paste the repo link into Claude with the words “install skill” and Claude installs it. You only need Python 3.8 or newer on your computer, and nothing else to install.
Can it place trades without me?
No. Only the broker skill can lead to an order, and only after the risk check passes and you type yes for that exact order. Paper trading is on by default, and the README says to keep order tools on “Needs approval”.
Does the Time Machine predict where the price will go?
No. It’s a simulation built from the stock’s own last 60 days of 5-minute bars. It gives odds, not a forecast. In the author’s Tesla example, it called the trade close to a coin flip.
Does it work for Indian stocks?
Not out of the box. The scanner uses US symbol lists from Nasdaq Trader, and the data comes from US sources. The method of planning a stop, a target and the odds first applies anywhere, but the tools would need changes.
Which skills should a beginner start with?
Start with the scanner, the exit planner and the Time Machine, on paper only. They teach you to decide your stop and target before you enter, and to check the odds first.
Want more free Claude setups? Read how I do SEO for a brand using 81 free AI skills, or how to make a motion graphics video with Claude AI. Got the skill running? Comment on the video and tell me which one you tried first.
Written by Gaznil, an AI-native digital marketer and SEO executive in Kochi, Kerala. I run SEO, social and paid ads for brands in India and the UAE, and I teach AI workflows to students and teams. I’m not a financial adviser, and this post is a software setup guide, not trading advice. Work with me.
Sources: Jake Schincariol, The Day Trader skill (README, SKILL.md files and config.example.json), GitHub, retrieved 2026-10-26. All figures for the Tesla example, the install steps and the PDT rule note come from this repository.
