Day Trader Skill is a free set of 11 Claude skills that scan charts, plan entries and exits and size risk for day trades, with paper trading by default.
What it is: a chain of small jobs, not one trading bot
The repo is by Jake Schincariol, MIT licensed, with no signup and no API key. Each skill does one job and leaves its result in a trader/ folder in your project, and the next skill reads it. The tools behind them are plain Python that read free public sources: Yahoo Finance charts, Nasdaq Trader's symbol lists, SEC EDGAR and the House Clerk.
/trader-scanner scans about 5,900 US stocks for setups such as opening-range breakouts, flags and oversold bounces./trader-news tracks headlines and filings on stocks you watch and tags what moves them./trader-entry watches a setup bar by bar and reports waiting, triggered or expired./trader-exit sets the stop and target before you enter, and a flat-by-close time./trader-timemachine replays the rest of the session 10,000 times and re-runs the rule on past days, covered below./trader-budget and /trader-risk size the position to your budget and maximum loss, then pass or block the order against your limits./trader-broker shows the order ticket and sends it only after you say yes. /trader-alerts rings on the stop, the target and the close./trader-analyst reads earnings reports from SEC filings. /trader-pelosi reads the stock-trade disclosures that members of Congress file.
One worked trade, as the README tells it
The README walks through Tesla on Monday October 19, 2026, with real data. At 10:15 AM ET the scan found 42 of 231 charts closed above their 15-minute opening range, and TSLA was one: range 354.47 to 362.18, with the 10:10 candle closing at 364.18. The exit skill put the stop at the range midpoint, 358.32, so 5.86 a share was at risk, with a target of one times that risk at 370.04.
- The Time Machine drew candles from TSLA's own previous 59 sessions: target first 29.4%, stop first 25.5%, still open at the close 45.1%.
- Re-running the same rule on those 59 real sessions, it fired 33 times, with 13 target hits, 15 stop hits and 5 neither. The README's verdict: close to a coin flip.
- With a $5,000 account and a $50 maximum loss, the budget skill sized 8 shares ($2,913.44), a loss of $46.88 if the stop fills at 358.32.
- The order skill then shows a plain-words ticket. You type yes for that exact order, and your brokerage connector, Alpaca's MCP server in paper mode by default, asks again through Claude's own Allow once or Deny card.
The safety design
- Nothing in the pack can place an order alone. Only
/trader-broker can lead to an order, and only after /trader-risk passes that exact order and you confirm it in the chat. "paper": true is the default in the example config. The README says to set your brokerage tool to "Needs approval", never "Always allow", and keep keys in environment variables, not in the chat.- No options, margin, short selling or leverage unless you turn them on.
- The README says the Time Machine is a simulation built from the stock's last 60 days of 5-minute bars. It cannot see an earnings report, a halt or a headline, so treat its odds as a reality check, not a promise.
Limits the author states
- It is not financial advice, and the README says most people who day trade lose money.
- A stop is a plan, not a guarantee. In a fast market or across a gap it can fill well past its price, and the maximum loss shown is the loss at the stop price.
- Yahoo Finance's chart endpoint is not an official API and can lag, change or rate-limit you. The scanner and entry timer read 1-minute or 5-minute bars, not every tick.
- The README says FINRA removed the $25,000 pattern-day-trader minimum on June 4, 2026 and that brokers may phase the change in until October 20, 2027. If yours still applies the old rule, set
"pdt_guard": true so the risk skill counts your day trades. Check this with your broker. - Members of Congress must disclose trades over $1,000 within 45 days, so the Pelosi skill reads public filings with a delay of up to that long, not inside information.
- Day-trading gains are short-term gains and wash sales are likely, and the README says nothing in it is tax advice.
Install and requirements
- In Claude, paste the repo URL and type "install skill". In Claude Code you can also run
/plugin marketplace add Jakeschincariol/day-trader-skill and /plugin install day-trader-skill@day-trader-skill. As a plugin the commands are namespaced, for example /day-trader-skill:trader-scanner. To get plain /trader-scanner, clone the repo and copy the trader-* folders into ~/.claude/skills/. - The tools need Python 3.8 or newer, with nothing to pip install. The SEC tools need your name and email in
trader/config.json or in the SEC_USER_AGENT environment variable. - Without Claude Code, you can paste a single
SKILL.md at the top of a chat and lose the Python tools. The skills run on a computer, not a phone.
Who is behind it, and how fresh it is
- The repository was created on October 20, 2026. On November 7 it had about 5,800 stars, 1,440 forks and 55 open issues, and its last push was the same minute it was created, so it has had no update since launch.
- The README calls it the sequel to the author's earlier Trader skill, which copies an investor's 13F, screens stocks and handles tax lots and rebalancing.
- The same author also made Instagram Agent Skill, a similar free set of Claude skills for writing Reels and captions.
- Third-party guides exist, including a setup guide by another creator and a separate India version called trader-skills-india, which uses NSE data. Those are not part of this repo.
- Seailife did not install it, run the scripts or place any order.