⬡ my-bot

Strategies, market data and a one-command Docker stack for Algorithex.

---

What this repo is

This is an Algorithex project workspace, not the trading framework itself.

It holds the parts that belong to one specific bot — your strategies, your stored market data, and the compose stack that runs them — so they stay version-controlled separately from the engine that executes them.

This repo — my-bot	Your strategies, your data, your settings
The framework — yashd-dev411/algorithex	The engine, dashboard and research modules

Clone this repo, point it at the published framework image, and you have a working trading stack. Nothing here needs to be edited to get started.

---

What's inside

Path	What it holds
strategies/	One directory per strategy, each with an __init__.py. The dashboard's Strategies tab discovers them from here
storage/	Generated candles, logs, charts and backtest results
docker/	The compose stack — app + PostgreSQL + Redis
.env.example	Every setting the stack reads, with the safe defaults
AGENTS.md	Rules for MCP-compatible AI assistants working on this project's strategies

---

Quick start

git clone https://github.com/yashd-dev411/my-bot.git my-bot
cd my-bot
cp .env.example .env

1. Point it at the published image

This step is easy to skip and hard to debug, so do it first.

The compose file declares a build context of ../../algorithex, which only resolves on a machine that already has the framework checked out beside this repo. A fresh clone has no such directory, so compose tries to build and fails with a missing-context error.

In .env, set the published image instead:

ALGORITHEX_IMAGE=ghcr.io/yashd-dev411/algorithex:latest

That image is public — no GitHub login is needed to pull it.

2. Start the stack

cd docker
docker compose up -d
docker compose ps

The dashboard is at http://localhost:9000. Sign in with the PASSWORD value from .env.

docker compose ps is worth watching on the first run. The app waits for PostgreSQL and Redis to report healthy before it starts, so you should see them pass and only then the app come up. The first boot is slow because it installs the workspace — the healthcheck allows 120 seconds for that before it starts counting failures.

---

The example strategy

strategies/ExampleStrategy/ is a working EMA trend-following starter. It goes long whenever the fast EMA is above the slow one, and sizes the position from a risk percentage of the account rather than a fixed quantity.

@property
@cached
def fast_ema(self):
    return ta.ema(self.candles, self.hp['fast_period'])

def should_long(self) -> bool:
    return self.fast_ema > self.slow_ema

It ships with five hyperparameters — fast_period, slow_period, risk_pct, take_profit_pct and stop_loss_pct — so you have something real to optimize in the dashboard rather than an empty framework to start from.

---

Configuration

Everything is set in .env at the repository root. Two values matter before anything can reach the dashboard:

Variable	Why
PASSWORD	The dashboard login. Its sha256 is what the API compares, so changing it invalidates existing sessions
POSTGRES_PASSWORD	Set this if the database is reachable from anywhere but this machine

The rest are local defaults, and the compose file falls back to the same defaults if a variable is unset.

---

Ports

Port	Service
9000	Dashboard
9001	Language server (editor intelligence)
9002	MCP server — lets an AI assistant drive this project
8888	Jupyter

---

Credits and License

Algorithex is a fork of Jesse by Jesse Mir and contributors, released under the MIT License. Upstream copyright is retained in LICENSE.

This workspace is distributed under the same terms. MIT permits use, modification and redistribution provided the original copyright and permission notices are kept intact — which is why they are still here.