HOW IT WORKS

Run it. Find the leak. Extend it.

Import your trades, see where your money goes in dollars, then open the full source and extend your own copy with your AI agent.

1 INSTALL

Unzip it — install one free thing

Unzip the download. TradeLens needs only Python installed (free, one time — on Windows tick “Add python.exe to PATH”). It sets up everything else by itself.

›_python.org · tick “Add to PATH”

TradeLens · unzippedyour files

📜 START-HERE.bat double-click · Windows

📜 start-here.command double-click · Mac

📄 START-HERE.md read me first

📁 app/ · core/ · connectors/ the source

📁 docs/ install · AI setup

🗃 demo-preview.sqlite3 426-trade demo

2 LAUNCH

Double-click the launcher

Double-click START-HERE.bat (Windows) or start-here.command (Mac). The first run sets itself up in about a minute, then opens TradeLens in your browser — no account, no login. Keep the little black window open while you use it.

›_opens http://127.0.0.1:8742

START-HERE.batfirst run

START-HERE.batFirst run - setting up TradeLens. This takes about a minute...Installing TradeLens into a private virtual environment (.venv)...TradeLens is installed.Starting TradeLens... browser opens at http://127.0.0.1:8742keep this window open · close it to stop

3 IMPORT

Load the demo, or import your trades

On the first screen, click Load demo data for a 426-trade sample to click through — or open Import and add your own MT4 / MT5 / CSV. The demo stays separate; Settings → Start fresh clears it anytime.

›_Load demo data · or Import MT4 / MT5 / CSV

TradeLens · Importlocal

4 REVIEW

Find the leak — in dollars

See your strongest edge, your #1 leak, and what your worst habit costs — each priced per month, computed from your real trades. The number you can’t argue with shows up first.

›_#1 leak · priced per month

TradeLens · Overview · DEMOlive

5 AI SETUP

Turn on the AI coach — free & offline

Optional. Run scripts/setup-local-ai.ps1 (Windows) or .sh (Mac/Linux) — it installs Ollama and pulls the bundled model (llama3.2:3b, ~2 GB once). The coach then runs 100% on your machine, no key. Prefer the sharpest? Paste your own Claude / OpenAI / Gemini key in Settings → AI Coach.

›_setup-local-ai.ps1 · or paste an API key

TradeLens · Settings · AI Coachoffline

TradeLens AI Coach settings - pick a model

6 ASK

Ask the coach — get exact numbers back

Ask your account anything in plain English. The local AI coach answers with figures TradeLens already computed — never a guess — and hands you a button straight to the trades behind them.

›_what should I stop trading this week?

TradeLens · AI Coachlocal model

7 EXTEND

Extend the source — with your AI agent

You own the full codebase. Open it in Claude Code or Cursor, hand over AGENTS.md, and add your own KPIs, connectors, reports or a prop-firm risk module. The 92 tests keep the numbers honest while you build.

›_add a prop-firm daily-loss risk view

TradeLens/ · your copyyours

📁 app/ FastAPI backend

📁 core/ analytics · 92 tests

📁 connectors/ MT4/5 · cTrader · CSV

📄 AGENTS.md AI-agent instructions

📄 START_HERE_AI.md hand to Claude / Cursor

Actual package · not a black box

You’re not buying a login. You’re buying the codebase.

Use it as a private trade desk today — or as a real codebase your AI agent can build on tomorrow. Not a template, not a prompt pack — a working local app with source, tests, docs and a demo database.

Python / FastAPIVanilla JS frontend92 testsAGENTS.md included

TradeLens/ · what’s in the box

app/the runnable local app (FastAPI)

core/analytics engine · deterministic calculations

connectors/MT4/5 · cTrader · TradingView · CSV import

web/frontend source · vanilla ES modules, no build

tests/92 passing tests

docs/install · how-to-use · AI-setup · extend

bonuses/prompt pack · templates · 30-day ritual · recipes

AGENTS.mdinstructions for AI coding agents

START_HERE_AI.mdhand this to Claude / Cursor / ChatGPT

demo-preview.sqlite3426-trade demo database

Give your AI agent a real codebase

Add a CSV template for my broker.

Create a prop-firm daily-loss risk view.

Add a KPI for trades taken after two losses.

Build a monthly PDF review report.

Audit the expectancy calc before changing it.

Use-only · no coding

Run the app. Import trades. Ask the coach. The whole desk works with zero code.

Build-on · with AI

Open the source, give AGENTS.md to Claude Code / Cursor, and extend your own copy.

In plain English

What you can ask TradeLens

No query language. Ask your account the way you’d ask a mentor — it answers from your real numbers.

What setup actually makes me money?

What should I stop trading this week?

What happens after I take a loss?

Which session should I avoid?

Which rule-break costs me the most?

What changed vs last month?

Full source · $47one time · Cloud sold separately

Get source