qianshe/dsh-assistant-optimization0

dsh-assistant-optimization

Fold mis-rendered thinking/reasoning content into collapsible blocks and render mermaid diagrams inline.

AI Analysis

核心用途是优化 DSH Web 的对话渲染体验。适合需要清晰查看模型推理过程、频繁交互 Mermaid 流程图的用户。

Package
dsh-assistant-optimization
Version
1.0.0
License
MIT
Last updated
Aug 18, 2026

Install

$npx -p @deepseek-ai/dsh dsh plugin --profile web add github:qianshe/dsh-assistant-optimization

Usage

After installation, the plugin works automatically — no extra steps needed.

  • Reasoning fold — When a model outputs `` (or any marker you configure), the reasoning text before it is collapsed into a "Think" disclosure. Click to expand/collapse.
  • Mermaid — Any ```mermaid code block in the conversation renders as an SVG diagram. Use the toolbar buttons (zoom in/out/reset), scroll wheel to zoom, or mouse/touch drag to pan.
  • Custom markers — Go to Settings → General → Thinking Tag Markers to add or remove split markers (different models may use different tags).

Configuration

SettingDefaultDescription
Markers["](https://github.com/qianshe/dsh-assistant-optimization/blob/84ccb69e7285130d2d6b1b37414c60f8a6dff42d/lib/client.js)Strings that split reasoning from body text. Text before a marker becomes a collapsible block. Multiple markers supported.

Configuration

SettingDefaultDescription
Markers["](https://github.com/qianshe/dsh-assistant-optimization/blob/84ccb69e7285130d2d6b1b37414c60f8a6dff42d/lib/client.js)Strings that split reasoning from body text. Text before a marker becomes a collapsible block. Multiple markers supported.