Day Trader Skill for Claude
Install Day Trader Skill, see how its tools work together, and understand the limits before you try a broker connection.
How to Install Day Trader Skill
Choose how you want to use the original Day Trader Skill GitHub project. Claude Code opens the full plugin installation; the manual route copies the skill folders on your computer. After installing, choose a Skill below to try a paper-trading task.
Claude Code — install from the original GitHub project
Source: github.com/Jakeschincariol/day-trader-skillFor Claude Code v2.1.275 or later. This is the author's GitHub project, not a skill from Claude's official marketplace.
Paste this into a Claude Code session
/plugin install day-trader-skill --marketplace https://github.com/Jakeschincariol/day-trader-skill.gitClaude Code will ask you to confirm the GitHub source and choose where to install the plugin. Finish those prompts there. This command installs the pack; it does not run a trading skill.
Verify the installation
After installation, open /plugin > Installed, or type / and look for /day-trader-skill:trader-risk. Seeing the name confirms the skill is available; it does not mean a risk check has run.
Using an older Claude Code version?
First add the GitHub source, then install the plugin. Complete each command separately.
/plugin marketplace add Jakeschincariol/day-trader-skill/plugin install day-trader-skill@day-trader-skillClaude Chat — understand the limitation
For this GitHub project's full plugin, use the Claude Code instructions in the first tab. In Claude Chat, you can discuss a pasted SKILL.md, but pasting that file does not give the chat access to the project's local Python tools. Open the GitHub project to read the source, or switch to Claude Code to install the full pack.
Manual setup — local Claude skill folders
Use a local terminal on macOS or Linux. You need Git, Python 3.8 or later, and a Claude environment that reads ~/.claude/skills/.
git clone https://github.com/Jakeschincariol/day-trader-skill.gitmkdir -p ~/.claude/skillscp -R day-trader-skill/trader-* ~/.claude/skills/If the clone reports that the folder already exists, inspect that folder before copying. After copying, check for /trader-risk or another plain skill name in your Claude environment. Run any local Python helper from the cloned repository root with its required trader/ files.
A project-local setup uses that project's .claude/skills/ location instead. To install one skill manually, copy only its corresponding trader-* folder.
Before running a local helper
The source README specifies Python 3.8 or newer and no additional pip dependency for basic scripts. Run local commands from the cloned repository root with the required trader/ inputs. SEC requests need your own name and email as SEC_USER_AGENT. Keep broker credentials in your own environment, outside this site or chat.
If installation seems to fail
Claude cannot find a skill? Confirm installation was approved, the plugin is enabled, the namespaced name is correct, and no reload is pending. A successful Copy is not installation.
If --marketplace is not recognized, check your Claude Code version and use the expanded two-command fallback. Missing Python files? Check the repository root and the Skill's required inputs. SEC data unavailable? Check your local user agent and service response.
What Is Day Trader Skill?
Day Trader Skill is Jake Schincariol's open-source collection of Claude skill instructions and local Python helpers for day-trading research and planning. Claude can use a skill's instructions to choose and explain a workflow; the Python helpers read market or filing data, write local files, and produce outputs for the next step. The package is not a standalone trading terminal, and this independent guide does not run its tools.
The current Day Trader Skill source package has 11 skills. Scanner and News help identify a research candidate. Entry and Exit put conditions around a possible trade. Time Machine compares simulated paths with past occurrences. Budget and Risk check proposed size and configured limits. Broker can show a ticket and, in a separately configured user environment, has a route that requires explicit approval before an order. Alerts, Analyst, and Pelosi add monitoring or public-data research. Each step requires the reader to inspect the result before continuing.
| Part of the package | What it does | What it does not establish |
|---|---|---|
| Claude skill instructions | Tell Claude when to use a named workflow and how to interpret its output. | A pasted instruction file alone does not run a Python helper. |
Local Python helpers and trader/ files |
Fetch or process data and pass saved results between steps in the user's own workspace. | A result is not a live order or a guaranteed market price. |
| Public data sources | Day Trader Skill uses Yahoo Finance charts, Nasdaq Trader symbol lists, SEC filings and House Clerk disclosures for different tasks. | Sources can be delayed, missing or revised; Yahoo's endpoint is unofficial. |
| Broker path | Can display a paper ticket; actual submission belongs to the user's configured broker environment and approval flow. | This website cannot connect to a broker, store keys or place a trade. |
The source README lists no paid market-data API key for the basic research tools. Claude plans, brokers and connected services set their own prices; check them before connecting an account. A scan match is a candidate to inspect. A simulation shows paths produced from past data. A Risk PASS reports which configured rules passed for the available inputs. None predicts a profit. Read the source repository · Read the Risk guide.
From a Stock Scan to a Paper Ticket
This Day Trader Skill walkthrough maps the source project's October 19, 2026 TSLA historical example. It shows what to inspect at each handoff. Readers using the cloned GitHub repository can follow the local paper-trading tutorial for commands and inputs. Claude Code plugin users can open the linked Skill guides for requests to send Claude.
| Stage | Skills used | Input and output to inspect | Decision before continuing |
|---|---|---|---|
| Find a candidate | Scanner → News | Scanner reports which charts met a rule and whether coverage was partial; News adds dated headlines and filings. In the Day Trader Skill source's historical scan, 42 opening-range matches came from 231 usable charts. | Verify the scan time, candidate symbol, source and news relevance. A hit is not a buy instruction. |
| Define the plan | Entry → Exit | Entry distinguishes a completed-bar trigger from WAITING or EXPIRED; Exit writes a planned stop, target and time boundary. |
Keep the exact entry, stop and target tied to the dated case; do not reuse historical prices today. |
| Test assumptions and size | Time Machine → Budget → Risk | Time Machine separates simulated paths from actual past rule occurrences. In the source example, 10,000 simulated paths and 33 historical occurrences have different denominators. Budget's illustrated $50 maximum planned loss and $5.86 planned risk per share limit the example to eight whole shares, or $46.88 at the planned stop. Risk returns rule-by-rule PASS or BLOCK. |
Review sample size, current input timestamps, cash and the reason for every rule result. PASS is not a safety certificate. |
| Inspect, do not submit | Broker | Trader Broker displays the saved paper ticket and readiness fields. | Compare ticker, quantity, mode, stop and target. This website stops at a read-only explanation and never sends an order. |
Alerts, Analyst and Pelosi support monitoring or separate research; they are not required for this example. Open the local tutorial · Read the original dated walkthrough.
Explore Skills
Choose the Day Trader Skill guide for the question you have now. The current package contains 11 skills. Every guide explains when to use its skill, a clearly labeled Claude prompt example, the output to read and a specific limitation.
Trader Scanner
Find charts matching a stated rule. Read scan coverage and PARTIAL before treating a symbol as a research candidate.
Trader News
Put headlines and SEC filings next to a candidate with their source and time. Separate an event from a trade decision.
Trader Entry
Check whether a completed bar triggered an entry condition, is still waiting or expired. A forming bar is not confirmation.
Trader Exit
Write a planned stop, target and time boundary before sizing. A stop price does not guarantee a fill.
Trader Time Machine
Compare simulated paths with separate historical occurrences. Read the sample and method behind each percentage.
Trader Budget
Convert a maximum planned loss, risk per share, cash and position limits into a whole-share proposal.
Trader Risk
Read why a proposed order passed or was blocked by configured rules, including the inputs and timestamps used.
Trader Broker
Inspect a paper ticket and the source-described approval path. This guide cannot approve or submit an order.
Trader Alerts
Set price or time reminders and check when they last ran. An alert is not guaranteed real-time protection.
Trader Analyst
Read filed SEC earnings facts, compare like periods and distinguish reported figures from derived values.
Trader Pelosi
Read public House trade disclosures with trade date, filing date, asset type and amount range kept separate.
Independent Checks and Community Reports
See what one outside offline test measured, then read two GitHub user reports before using a broker connection. These sources do not establish live trading performance.
Independent offline check — Nanoflow. The author reports that Day Trader Skill's 315 offline tests passed in their environment. They also compared simulated and observed target-first outcomes across 428 historical setups: 38.9% in the simulation versus 33.9% in the observed sample. This supports a limited account of saved-data behavior under that test method. It does not show that this site tested the code, or that a future trade is profitable. Read the independent test.
Reported price-side risk — GitHub Issue #32. A user reported a market-order risk check passing after a fresh price had crossed a planned stop or target. The issue was open when checked on November 5, 2026. It is a specific community report, not proof that every order is affected or that a fix has shipped. Read Issue #32.
Reported settled-trade simulation case — GitHub Issue #19. A user reported that Time Machine could still display simulated odds when available past bars had already settled a trade. The issue was open when checked on November 5, 2026. Read the original case and current source status before relying on the tool's interpretation. Read Issue #19.
Read the full review for how the source documentation, outside check and issue reports fit together.
Guides & Tutorials
Choose the source-linked review to assess the full pack, or use the local Python tutorial to follow its dated paper-trading example. Both articles link to the original project and the relevant Skill guides.
Day Trader Skill Review: What the 11-Skill Toolkit Actually Does
Read one outside offline test and two open GitHub issues, then see what to check before using the tools.
Read reviewTutorialHow to Use Day Trader Skill: A Local Paper-Trading Tutorial
Clone the GitHub repository and run its Python tools from scan to a read-only paper ticket. Claude Code plugin users can follow the Skill guides.
Read tutorialFrequently Asked Questions
What is Day Trader Skill?
It is an open-source set of Claude skill instructions and local Python helpers for day-trading research and planning. Its current source package contains 11 skills. TraderSkillsHub is an independent guide to that project, not a broker or the project owner.
Is Day Trader Skill free?
The Day Trader Skill source project is MIT licensed and free to obtain. Claude, brokers and connected services set their own prices; check their terms before using paid features.
How do I install Day Trader Skill in Claude?
In Claude Code v2.1.275 or later, copy the GitHub-source plugin command in How to Install, then confirm the source and install scope in Claude Code. Older Claude Code versions have a two-command fallback. Pasting a Skill file into Claude Chat does not install its local Python tools.
Can I install the whole pack at once?
Yes. The Day Trader Skill plugin and full manual-copy routes include the current skill folders together. You do not need to reinstall a Skill from its Wiki page after the full pack is present.
Can I install only one Skill?
Yes, with a manual folder copy. Select the corresponding trader-* folder and read its Skill guide for prerequisites. Some tools need files produced by earlier steps.
Why can Claude not find a Skill after I copy it?
Copying on this site only puts text on your clipboard. Complete Claude Code's installation prompts, check the plugin in /plugin > Installed, and follow any reload message. Plugin skills use the /day-trader-skill: prefix; manually copied folders use plain names.
Can this website trade automatically?
No. This site publishes Day Trader Skill guides and text to copy. The source project describes a separate broker path in the user's environment with risk checks and explicit approval. Read the Broker guide before considering it.
Does a Risk `PASS` mean a trade is safe?
No. It means configured checks passed with the available inputs. Data can be stale, and actual execution can differ from a planned price. Read the Risk guide and the linked issue report.
Does Time Machine predict what a stock will do?
No. Day Trader Skill's Time Machine simulates paths from past bars and checks past occurrences separately. Its percentages have different methods and sample sizes and cannot predict a future outcome. See the Time Machine guide.
Where should a beginner start?
Install the GitHub plugin in Claude Code, then open the Trader Scanner guide and use its sample request with a paper task. Inspect each result before moving to the next Skill. If you cloned the repository and use a terminal, follow the local tutorial. The dated TSLA example is not a current market signal.