Independent software studio

Melogne Studio

Building lightweight native Windows applications.

A small studio focused on fast, focused desktop software — built with Rust, Tauri, React and a deep respect for modern Windows.

View Projects GitHub

// apps

Projects

Native Windows applications, crafted with care.

Available

Lightweight screen-time tracker for Windows. Tracks app usage, session snapshots, and daily stats — built with Tauri 2, React, Rust and SQLite.

Tauri 2 Rust React SQLite Windows

GitHub Releases

TrackMLN screenshot

Minimal encrypted password vault with a clean desktop UI. Simple, secure, and gets the job done — complete and archived.

Python CustomTkinter AES Encryption Windows

GitHub Releases

VaultMLN screenshot

In Development

A fast, local-first algorithmic trading platform built in Rust. Engine-first — UI follows.

Rust Tauri 2 React WASM Rhai SQLite Dhan API WebSockets

This project is not yet released.

// about

About

The person behind the studio.

WTRMLN

WTRMLN

Middle school developer

Middle schooler building desktop apps. Focused on Tauri 2, Rust and Win32 experimentation — desktop software, overlays, and low-level Windows tooling. Independent software studio, all projects live on the GitHub org.

// Stack

Rust TypeScript React Tauri 2 SQLite Win32 API DLL Injection Python Windows

// github

Organisation

Find us on GitHub.

Organisation

Independent software studio. Building lightweight Windows desktop apps — screen-time trackers, password vaults, and more.

Visit on GitHub

Members

Middle schooler. Main developer @ MelogneStudio.

Visit on GitHub

App tester.

Visit on GitHub

// Overview

A fast, local-first algorithmic trading platform built in Rust. Engine-first, not UI-first — every layer is tested, deterministic, and production-grade before the next layer is added.

One execution engine drives backtests, paper trading, and live trading. No cloud dependency, no mandatory account, paper trading is always the default.

// data flow

Data → Indicators → Strategy Engine → Backtesting → Execution → UI

// Status

Phase 1 — Data: Broker trait, DhanClient, 1,000-symbol WebSocket manager with auto-reconnect, tick fan-out, historical OHLCV fetch.
Phase 2 — Indicators: MA, EMA, RSI, ATR, VWAP, Relative Volume, Bollinger Bands. Pure functions, stateless.
Phase 2.5–2.9 — Strategy Engine: Custom DSL (lexer → parser → AST → validator → runtime), trigger state, cross detection, paper broker, deterministic candle-by-candle replay.
Phase 3–5 — UI: Visual strategy builder, strategies list, coder modal, uploader modal, settings, browser-only fallback.
Phase 6 — Plugin System: Rhai + WASM runtimes, capability gating, broadcast event bus, desktop Plugins screen.
Phase 7 — Live Trading: pending — broker-agnostic ExecutionTarget facade, two-step confirmation, risk limits, immutable trade log.

132 passing · 1 ignored · 0 failed (run cargo test --workspace)

// The Strategy Language

Strategies are written in .algomln files. The language is intentionally small — rules only, no variables, no loops.

// grammar.algomln

strategy       = rule+
rule           = "WHEN" condition NEWLINE action

condition      = comparison
               | cross_expr
               | not_expr
               | logical_expr

comparison     = expr operator expr
operator       = "<" | ">" | "<=" | ">=" | "==" | "!="

logical_expr   = condition "AND" condition
               | condition "OR" condition

not_expr       = "NOT" "(" condition ")"

cross_expr     = "cross_above" "(" expr "," expr ")"
               | "cross_below" "(" expr "," expr ")"

expr           = indicator_call | price_field | number
indicator_call = indicator "(" integer ")"
indicator      = "ema" | "ma" | "rsi" | "rel_vol" | "atr" | "vwap"
               | "bb_upper" | "bb_lower" | "bb_mid"

price_field    = "close" | "open" | "high" | "low" | "volume"
               | "prev_close" | "prev_open" | "prev_high" | "prev_low"

action         = "BUY" integer
               | "SELL" integer
               | "SELL" "ALL"

Blank lines and # comments are allowed anywhere. Keywords are case-insensitive.

// Examples

// rsi.algomln — oversold / overbought

WHEN rsi(14) < 30
BUY 1

WHEN rsi(14) > 70
SELL ALL

// ema_cross.algomln — EMA crossover

WHEN cross_above(ema(20), ema(50))
BUY 10

WHEN cross_below(ema(20), ema(50))
SELL ALL

// bb_breakout.algomln — compound + Bollinger Band

WHEN ema(9) > ema(21) AND rsi(14) < 60
BUY 5

WHEN close > bb_upper(20)
SELL ALL

// Performance

Backtest replay on 184,863 candles (full NIFTY 1-min history): 3.5s · 52,000 candles/sec · 9,026 trades. Same strategy + same candles = identical results every run — no randomness, BTreeMap over HashMap in any iteration path that could affect output.

// Plugin System

Plugins load from <app_data>/plugins/<id>/ and run in a Rhai or WASM sandbox with capability gating (Market Data, Storage, Indicators, Analytics, DSL Extension, UI Panels, Scheduler, Execution). Backtests never wire the event bus, so plugin callbacks can't run during replay — this keeps backtests deterministic.

// Roadmap

Position sizing rules, stop loss / take profit, risk controls, multi-symbol strategies (Phase 6.5)
Live broker execution via ExecutionTarget with two hard confirmation steps (Phase 7)
Immutable trade log, user-defined risk limits (max loss, max orders)
Risk acknowledgment on first live trade

// Run It

// cargo

cargo run --release --bin behavioral_backtest -- run my_strategy.algomln \
  --data sample-data/nifty_1min.csv --symbol NIFTY

More options: --candles 10000 · --cash 500000 · profile rsi 50000 · profile ema

// Links

// Screenshots