Simple Chart

Simple Chart is a desktop stock charting application for swing traders that use the AVWAP.

This repository is structured for source-based local installation on Linux, macOS, and Windows. The intended flow is:

1. Clone the repo.
2. Create a Python 3.13 virtual environment.
3. Install with pip install -e ..
4. Launch with simplechart.

No hard-coded personal paths are required.

Requirements

• Python 3.13
• pip
• A desktop environment capable of running PyQt6 applications
• Internet access if dependencies need to be downloaded during install

Fast Path

Linux and macOS

python3.13 -m venv .venv
. .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -e .
simplechart

If python3.13 is not available, use whichever command on the machine resolves to Python 3.13 and verify with:

python --version

Windows PowerShell

py -3.13 -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install -e .
simplechart

If py -3.13 is unavailable, install Python 3.13 first or use an equivalent launcher that targets Python 3.13.

What Gets Installed

pip install -e . installs:

• Python dependencies from pyproject.toml
• the simplechart command defined by the project entry point

The application stores its local SQLite database in a per-user location by default:

• Linux and macOS: ~/.simplechart/simplechart.db
• Windows: %USERPROFILE%\.simplechart\simplechart.db

You can override that location:

simplechart --db /path/to/simplechart.db

Linux Desktop Launcher

Running from a terminal after pip install -e . is enough to use the app.

If you also want a desktop launcher and icon in Linux application menus:

python scripts/install_linux_desktop.py

That script installs:

• io.simplechart.SimpleChart.desktop
• assets/simple-chart.svg

into the current user's XDG data directories. It also rewrites the installed desktop file so Exec and TryExec point at the simplechart executable from the environment where the script was run.

Platform Notes

Linux

• Use the fast path above.
• If a launcher icon does not appear immediately after installing the desktop entry, log out and back in or restart the shell session for the desktop.

macOS

• Use the fast path above.
• This repo currently targets source-based execution, not a native .app bundle.
• If a future release needs a native app bundle, use PyInstaller or py2app.

Windows

• Use the PowerShell fast path above.
• This repo currently targets source-based execution, not a native installer.
• If a future release needs a native installer or shortcut setup, use PyInstaller or another Windows packaging tool.

Troubleshooting

simplechart command not found

The virtual environment is probably not active, or the editable install did not complete successfully. Activate the venv and rerun:

python -m pip install -e .

Wrong Python version

This project requires Python 3.13. Check with:

python --version

GUI does not launch

Common causes:

• running in a headless environment
• missing desktop GUI support
• dependency install failure earlier in the setup flow

Try launching from the same terminal where the install was performed so any Python traceback remains visible.

Building Custom Indicators

SimpleChart has a stable public API for indicator authors. Every import your indicator needs comes from one place:

from simplechart.api import Indicator, register, OHLCVSeries, bars_for_n_days
# ... and any other names you need

Indicators come in two kinds:

• Chart indicators — plot on the price chart, sharing the price y-axis (SMA, EMA, AVWAP). These are the default.
• Panel indicators — draw in a dedicated panel below the chart with their own y-axis (RSI, MACD). Override render_target() to return a short lowercase string naming the panel (e.g. "rsi").

Drop a .py file (or a directory with __init__.py) into indicators/ and it loads automatically on next launch — no manual wiring required.

A skill file walks an AI agent through the full process — reading the right reference code, deciding whether a compiled kernel is needed, implementing each piece in order, and verifying the result.

The skill file is located at:

Agent	Skill location
Claude	.claude/skills/new-indicator.md
Codex	.codex/skills/new-indicator.md
Gemini	.gemini/skills/new-indicator.md
Grok	.grok/skills/new-indicator.md

To use it, tell your agent: "Use the new-indicator skill."
The agent will read the skill file, ask you a few questions about what the indicator should do, and guide you through building it step by step.

A note on intraday SMA and EMA warmup

A 200-day SMA needs 200 trading days of bar data to produce its first valid value. On intraday timeframes (5m, 15m, etc.) that means thousands of bars. Free data sources such as yfinance only provide roughly 60 days of intraday history, which means the 200-day SMA would show NaN (blank) for the entire chart on intraday views.

To work around this, SMA and EMA use a warmup fill: the controller passes the full daily bar history as _daily_bars in the indicator params. For any intraday bar that falls in the NaN warmup zone, the indicator substitutes the daily SMA/EMA value for that trading day. The result is a continuous line on intraday charts even when the intraday history is too short to compute the full warmup from scratch.

With a subscription data source that provides extended intraday history (e.g. Alpaca with several years of minute data), the warmup zone shrinks or disappears entirely — the intraday bars alone are sufficient to compute the full period. In that case _fill_warmup_from_daily still runs but finds nothing to fill because there are no NaN values left in the warmup range.

New indicators that use day-based periods and are expected to behave correctly on intraday charts should follow the same pattern. See indicators/sma.py for the implementation.

Notes for LLM-Assisted Setup

An agent installing this repository should:

1. Read README.md.
2. Read pyproject.toml.
3. Create a Python 3.13 virtual environment.
4. Run python -m pip install -e ..
5. Launch simplechart.
6. On Linux only, optionally run python scripts/install_linux_desktop.py.

Constraints for an agent:

• Prefer the documented source-install flow over introducing PyInstaller or other packaging changes.
• Do not rewrite project structure unless setup fails for a concrete, platform-specific reason.
• Make the minimum necessary platform-specific adjustments and explain them.