AI Agent Hub
Back to plugins
🧰

dsh-stock-chart

Web Tools Updated 2026.08.29

Run the following command in DeepSeek Harness:

dsh plugin install xlin20021/dsh-stock-chart

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install xlin20021/dsh-stock-chart to install this plugin. Source: https://github.com/xlin20021/dsh-stock-chart

About this plugin

When working with stock technical analysis in DeepSeek Harness, you often need to revisit candlestick charts, moving averages, MACD, and KDJ indicators repeatedly. Traditional approaches either force the model to describe everything in text (token-hungry and hard to scan) or store charts in browser localStorage that vanishes on refresh or in a new tab. dsh-stock-chart is built for exactly this gap.

The plugin renders interactive daily and weekly candlestick charts with TradingView Lightweight Charts, automatically overlays MA5/10/20/60/120, Bollinger Bands, MACD, KDJ, and volume indicators, and includes a drawing toolbar for lines, rectangles, and text annotations. All data is written directly to a local SQLite database (~/.dsh/stkdata) via stk-fetch.mjs without passing through the model context, which saves tokens and ensures charts survive restarts, browser switches, and even device changes.

It is ideal for users who track A-share technical indicators inside DeepSeek Harness and need to pull up historical candlestick charts quickly during a conversation. Simply provide a stock symbol and the plugin guides through fetching, persisting, and rendering the full chart.

Use Cases

  • Render daily or weekly candlestick charts in-conversation with MA, BOLL, MACD, and KDJ overlays
  • Restore historical charts from SQLite after a restart or browser switch without re-fetching data
  • Annotate key support and resistance levels with lines, rectangles, and text on the chart

Best For

  • Investors tracking A-share technical indicators inside DeepSeek Harness
  • Technical-analysis enthusiasts who review and annotate historical K-line charts in a conversation
  • Harness users who want data fetching bypassed from model context to save tokens