Plugins

DeepSeek Harness plugin: secondary working directories for a project. The agent keeps the primary workspace as cwd, gains equal write/exec permissions on configured secondary directories under workspace-write mode, and is notified of configuration changes

分流对话（Plugins of Conversation Departure）：为 @deepseek-ai/dsh（DeepSeek Harness）提供的侧边栏插件，把项目上下文打包成结构化 prompt 并复制到剪贴板，配合同屏打开的 DeepSeek 网页版做项目阅读/答疑；支持 PDF/Word/PPT/Excel 文本提取与图片 OCR，自动发现本机 poppler。纯本地：不联网、不调网页接口、不操控浏览器。安装：dsh plugin add dsh-dual-bridge dsh-bett

Ctrl+Space command palette over the Harness Web UI: one box to find sessions, invoke skills, open panels and switch appearance.

为 dsh Web 侧边栏提供两级分组：工作区分组（把一组工作区打包管理，只出现在列表最外层）与会话分组（每个工作区内部的会话归类）。

由 AI 协作的个人学习型知识梳理工作台

Ranks DSH plugins by real monthly npm downloads instead of GitHub stars, and names the plugins that star ordering under-rates. It reports the disagreement between the two orderings rather than a single quality verdict, and every number is a live read from npm's public downloads endpoint at call time, not a stored snapshot. A second dimension answers what downloads cannot: it reads the tools version line a plugin declares in peerDependencies and compares it with the version running here, so a popular plugin that cannot be installed on this host is reported instead of recommended. Read-only GETs, no HTTP route of its own.

Read-only code-reading workspace with multi-repository projects, source-linked notes, code maps and Mermaid diagrams.

Connects DeepSeek Harness to MCPTender public Russian procurement records, source documents and published contract suppliers using a personal service API key.

Connects DeepSeek Harness to MCP Bids public US federal opportunities and connected TED notices, with source-linked document tools and personal API-key authentication.

Helps keeping tracked updates from plugins added from github with a simple usability UI.

Prefixes every session title with the name of its workspace or project, so a flat session list shows which project each conversation belongs to; titles you renamed yourself are left unchanged.

Chat quotes for DSH: select text in any message to quote or comment; the quote waits in a Chat quotes chip on the composer (never inside the draft), expands on hover into selected text and your comment, and folds into the message only when you send.

Agent preset with adaptive tool presentation and environment-aware shell routing. Ports modules from yjh051108/dsh-routing-suite and xiaobright/dsh-anchored-standard (the former is listed here), so do not enable it alongside those presets. Requires DSH 0.2.2 or later; high uses PTC with on-demand tool unlocks and max uses PTC with the full available tool catalog.

Kiro (AWS CodeWhisperer/Amazon Q) provider route for DeepSeek Harness, plus an OpenAI/Anthropic reverse proxy. Reuses the existing Kiro IDE/CLI sign-in.

Plugin Steward adds a plugin panel to the chat composer for managing the installed DeepSeek Harness plugins: list them, switch them on or off, edit a plugin's settings, hide one plugin's tools or MCP servers from a single conversation, and have the model write short Chinese labels for installed plugins by reading their own README and package files.

Read-only DSH profile doctor: the seven config-tree checks (config-clobber, dead-patch, tool-collision, plugin-not-mounted, plugin-stale, entry diff, explain) plus eight ecosystem checks learned from real failures — pnpm store version conflicts that break `dsh plugin add`, patch insert names node_modules cannot resolve, broken junctions, duplicate-capability plugins, orphan backup dirs, GitHub-source deps, npm publish-scope hints, and the minimumReleaseAgeExclude gap. Four agent tools; every finding carries its fix.

Sends desktop notifications and custom HTTP pushes when tasks finish or need your attention.

System-1 decision guard for DeepSeek Harness, 4 questions asked before tool calling (risk reversible, relevance, injection suspension) and allow/ask/reject action. Needs an API key for a system-one compatible provider; decision requests are sent there.

Syntax-highlighted code for reports via one highlight_code tool: 68 languages, 15 themes, 11 output formats (inline HTML, Word-ready SVG/PNG, LaTeX), incl. the classic SyntaxHighlighter themes. PNG output needs Python 3 with Pillow; other formats need no external runtime.

Drives the DSH right-sidebar built-in browser from the agent side with 38 tools (open, navigate, read page text and interactive elements, accessibility tree, click, fill, select, check, key press, scroll, resize the browser frame by dragging its corner, with the page inside reflowing responsively so a narrowed frame shows the mobile layout, screenshot, full-page screenshot, PDF, markdown export, scrape to JSON, download, cookies, per-step operation records with replay and action whitelisting), serves the built-in web_search from a self-hosted SearXNG over an SSH tunnel with Tavily and a real-browser fallback, and registers a fetch provider that retries blocked GitHub URLs through measured mirrors and sniffs charset from the response header so GBK and GB2312 pages decode correctly.

Nine model-facing tools over the qBittorrent WebAPI v2, with confirm-gated mutations and an orphan-torrent diagnostic.

Park whole tool packs out of the session's tool surface and mount them back with one command. Cuts the tool table from 99,789 to 48,144 characters (51.8%) without uninstalling plugins or restarting. Reversible per group, per agent.

Adds a manual "mark as unread" action to workspace session rows, and retires itself once the harness exposes the same capability.

Tree-style questioning for DSH. Hover the "Question tree" tab above and a small card pops up with the whole tree — identical to the sidebar, but it frees that column and is quicker to consult (click "List / Graph" in the card to switch views). Select text in a conversation, click "Ask in tree", and a node appears in a tree you can reach three ways — hover the "Question tree" tab for a floating card that leaves the conversation in place, open the same tree as a full conversation view, or pin it in the right sidebar. Each node holds the quoted passage plus your own questions; clicking another node makes it the current parent so later selections nest under it (Alt+click opens a root node, Ctrl+click appends another passage to the same node). Double-click jumps back to the quoted text and pages older history in automatically when the passage is not loaded yet. Hover a truncated node to read the full quote, switch between a list and a horizontal branch graph (the graph scrolls sideways with the wheel as depth grows), mark nodes understood with a red/green dot — or let "mark understood once answered" do it when the reply lands. Nodes live in localStorage; nothing is injected into the prompt.