Plugins

Task reliability: auto-retry on model timeouts/failures, auto-continue unfinished tasks, completion verification, thinking-repeat interruption, resume after sleep or restart, autopilot mode and a remote trigger endpoint (npm: `dsh-task-reliability`).

A mid-run steer input for DeepSeek Harness Web: while the agent works, an input bar appears above the composer to inject modification requests into the next decision via agent.steer(), without stopping the conversation. Drafts persist per session, unread injected messages can be recalled from the bubble, and an optional guiding prefix is supported.

Multi-platform API balance and usage dashboard for the DSH web GUI: balances for DeepSeek, Zhipu GLM, Kimi, StepFun, SiliconFlow, MiniMax, OpenRouter, Novita and xAI under the composer, estimated per-session and per-subagent cost with DeepSeek peak/off-peak pricing, automatic discovery of relays configured in DSH settings, and an optional draggable whale widget.

Thinking in Chinese: injects a system prompt so thoughts and replies stay in Chinese, expands the reasoning block by default with interactive collapse, and localizes UI labels (npm: `dsh-think-zh-expand`).

Access DSH from your phone or anywhere: Cloudflare Quick Tunnel public link + LAN HTTP/HTTPS direct access (auto self-signed cert), token auth, gzip compression, floating phone-icon panel with QR codes, and a NapCat QQ bot channel to fetch the link.

Agent governance kingdom for DeepSeek Harness: territories, role bindings (bind/unbind/session with model/agent identity), and a plan→assign→execute→review task loop where Worker claims never auto-complete (Claim ≠ Fact), with a local GUI operations console.

Gives text-only models vision: forwards user images to an OpenAI-compatible vision model and shows the descriptions in a Web UI right panel.

DSH Web UI 插件应用商店：目录浏览/搜索/一键安装/兼容徽章/已装列表。A plugin store for DSH Web UI — catalog browse, search, one-click install, compat badges, installed list.

DeepSeek brain + automatic image transcription: attach images in the GUI and each one is transcribed via the official deepseek-v4-flash-vision-exp by default (a pure-text V4-Pro brain can see images), with any OpenAI-compatible VLM or local Ollama as alternatives.

Windows desktop vision and control for any application that has a window — a native editor, Excel, a proprietary installer, a game launcher — rather than a browser: screen_view returns the live screen as a real image block the model can see (never OCR or a text summary), desktop_control drives the pointer, keyboard and windows inside that same per-monitor-DPI-aware pixel space, and desktop_sequence runs a whole batch of actions in one call — bounded by maxSequenceSteps (24 by default) — so a task costs one round trip and one frame instead of one of each per action.

DeepSeek Harness 默认 System Prompt 管理插件：在 Web GUI「设置」里浏览、检索并覆盖最终装配进每轮会话的系统提示词。

DeepSeek Harness bundle registering one skill that turns long-form production (academic papers, industry analysis, business commentary) into a 9-role pipeline: T1 literature / T2 data / T3 case scouts running in parallel, T4 analyst, T5 writer, T6 critical companion, T7 auditor, T8 finalizer run by the coordinator, T9 peer reviewer — across 6 phases with triangular evidence verification, a 23-check M-gate, a G0-G14 audit including a Chinese AI-trace gate, peer-review scoring with journal matching, and 4 human checkpoints. Installs with `dsh plugin add lunheng-article-pipeline` (package.json#dsh.bundle.patch plus a lib/index.js entry that registers the skill); ships role cards for T1-T9, templates and 11 zero-dependency verification scripts, with five-language READMEs.

Installs Herta as a DeepSeek Harness agent: a preset carrying her persona and memory prompt sections, six memory and voice tools, a narrative layer that separates her thinking from what she says, 80 bundled voice clips, offline and MiniMax speech synthesis, and two conversation views that render the session with her own components.

DeepSeek Harness plugin that turns Jev judgments into three automatic actions — pruning oversized tool output to the segments relevant to the request, screening fetched pages for instructions aimed at the model, and picking which skill fits the next step — plus the jev_ask and jev_gate tools, a judgment ledger that survives restarts, and a /jev-status report. Needs a TypeSafe API key; without one it is inert and sends nothing.

Bridge Claude Code's .claude/ directories into DSH with zero migration: project and user-level skills, commands, rules (one system-reminder envelope), agent files, hooks (PreToolUse/PostToolUse/UserPromptSubmit/ SessionStart/SessionEnd) and .mcp.json servers, plus Claude Code plugin marketplaces. Ships /cc-plugin for full plugin/marketplace management (claude CLI first, self-implemented fallback), /cc-resume to list and import Claude Code sessions into DSH as native event-sourced logs, and /cc-export to write DSH skills back as Claude Code skills. Two-way Claude Code <-> DSH bridge, versioned and released as dsh-claude-compat.

DSH Prompt Library plugin: prompt CRUD, tags & categories, import/export, AI polish, one-click clip selected text, sidebar and assistant bubble reminders.

The most complete GitHub integration for dsh: repo overview, releases, issues, pull requests, contributors, search, and a deep repo report with TTL caching — no API key needed.

Per-event sound notifications: turn completion, approval, question, plan-review, goal-blocked, and task-failure each get their own sound and volume, configurable in the Web UI (built-in synth, mute, or local audio file).

Route eligible shell commands through rtk to compress tool output and save tokens.

A simple DSH queue panel for editing, removing, steering, reordering, and bulk removal.

Runtime-switchable model output styles with Claude Code outputStyles parity, plus the output.render.* presentation protocol: a /style command, per-session persistence, systemPrompt injection, six built-in styles, a web picker, and a renderer registry with per-session/per-tool rules and /export.

Multi-platform research and deep search pipeline: harvest_scout for parallel discovery across GitHub, Web, Twitter, Reddit, Xiaohongshu, Bilibili, and YouTube, harvest_deep_research for automated comprehensive research reports, plus extract, cross-source verification, and credibility auditing.

Auto-fills reasoning-efforts, context windows, output limits and image modality fields for models from unofficial (custom) providers (with data from models.dev), and provides user experience improvements such as compatibility enhancement and remembering reasoning effort.

An AI routing gateway that runs as a DSH plugin. Adds a Routing System sidebar panel for suppliers, account pool, fallback combos and API keys, and serves an OpenAI-compatible endpoint at http://localhost:3080/v1.