dsh-quant-workspace

Quant research often means bouncing between data scripts, backtesting frameworks, and plotting libraries, with a pile of glue code in between. dsh-quant-workspace collapses that pipeline into a single DeepSeek Harness conversation: describe what you want in plain language, and the bundled Python engine pulls Yahoo Finance daily bars, runs a rule-based backtest, and emits a self-contained interactive HTML report — all without leaving the chat window.

Backtest output includes a per-trade table, total return, max drawdown, win rate, average holding period, and a buy-and-hold baseline. The interactive charts carry candlesticks with Bollinger bands and entry/exit markers, volume, %B, and an equity curve, with cursor-anchored zoom, pan, crosshair, and a colour-scheme toggle — a single HTML file with zero external dependencies. Strategies you have researched can be saved to a built-in registry and reused by ID; a Bollinger mean-reversion example rule ships out of the box so the workspace works immediately.

The plugin is strictly read-only: it places no orders, modifies no positions, and requires no market-data API key. It suits quant enthusiasts, strategy researchers, or any developer who wants to close the idea-to-validation-to-visualisation loop inside a chat interface without assembling a separate project pipeline.