Plugins

把长文本渲染成图片发送，利用每图 384 token 封顶压缩 LLM 输入 token，专为DeepSeek Harness设计的插件；Pack long text into images to cut LLM input tokens (384/image cap) — a DeepSeek Harness plugin.

The small deterministic operations an agent needs mid-task. JSON/YAML round-trip, JSON Schema validation, text diff, JWT decode, regex testing, SQL dialect transpiling, QR codes, timezone and unit conversion, page screenshots.

88API Image Studio for DSH: four Image2 and Nano Banana models for text-to-image, multi-reference editing, 2K/4K output, and sequential batches.

DeepSeek Harness 的 MCP 服务管理器：在 设置 → 插件 → MCP 标签页中发现 GitHub 上的 MCP 服务仓库（topic:mcp-server / modelcontextprotocol），探测 npm / PyPI 发布状态与 README 启动线索，详情查看后一键安装 / 移除。服务以 @deepseek-ai/dsh-mcp-client 行登记进 cordis.patch.yml，HMR 热重载，无需重启。支持 GitHub 代理。

Firefox browser control for DeepSeek Harness: token-authenticated WebSocket bridge + text-first browser_* tools driving the user's own Firefox via the companion AMO extension

DSH web plugin: Codex-style /side side conversations. Open a temporary fork of the current chat in a floating panel — ask side questions without interrupting the main task or polluting its context.

Context7 MCP bridge: gives the model up-to-date, version-specific library documentation and code examples through the Context7 MCP server, keyless by default.

Windows minimal agent preset for DeepSeek Harness: fixed one-line persona, three tools, no runtime context.

Game-style floating HUD: balance HP bar, context MP bar, official peak/valley pricing with countdown, auto-compaction under 5% remaining, and memory-carrying new conversation.

Text viewer and editor for the dock workbench: undo/redo, Ctrl+S save, unsaved-change confirmation and binary detection, registered as the default file viewer.

GLM 视觉模型插件：注册 glm-vision 供应商路由（glm-4.6v 系列，image+text 输入声明）并提供 glm_vision 工具，让 DeepSeek 等文本主模型直接调用智谱视觉模型看图。

DeepSeek Harness plugin: a vision worker over Cloudflare Workers AI (@cf/moonshotai/kimi-k2.6) that routes image requests from text-only callers, returns a versioned righthand.vision.v1 envelope, and supports follow-up questions.

China-ready voice input for the composer. Requires a local Python bridge (pip install dashscope websockets, run bridge/voice-bridge.py) — the plugin alone does not work. Browser mic streams 16 kHz PCM to the bridge, which runs Alibaba Cloud DashScope ASR (paraformer-realtime-v2); interim text fills the draft at the cursor, silence auto-stop, optional auto-send.

Screen capture and external vision recognition: take_screenshot, list_windows, analyze_image and view_image tools with a configurable GPT vision channel (gpt-5.5 / gpt-5.6-sol / gpt-5.6-terra), API key via the credentials service, and a settings card; view_image shows the screenshot in the Web UI while the model context keeps text only.

LongCat-2.0 LLM adapter: 1M context, binary thinking mode, tool calling, and streaming over the OpenAI-compatible endpoint.

Workspace/chat context menu for the DSH home page: pin, rename, open in file explorer, archive, fork, copy, and open in a new window.

Unified appearance coordinator: theme color presets, UI fonts, UI font size, plus per-workspace markdown font size / line-height / background, with a coordination layer that stops UI scaling from leaking into chat body text.

Browser Web Speech dictation for the Composer: recognition needs no dedicated ASR server, key, or model download; reuses Session text and a configured DSH model for contextual phrase hints and optional transcript polishing.

Edit your `~/.dsh/AGENTS.md` global rules from the Settings page: a text editor with save button, no command line needed.

Text-to-speech for DeepSeek Harness: speak agent replies in the Web UI with a provider fallback chain (OpenAI, ElevenLabs, Google, Azure, Groq, Deepgram, OpenRouter, Edge, Piper, eSpeak).

领域建模：主动构建并打磨项目的领域模型（CONTEXT.md 术语表 + ADR 决策记录），挑战模糊术语、推敲边界场景。受 mattpocock/skills（223k★ MIT）启发。

Let text-only models see images in DeepSeek Harness: intercepts the llm/stream waterfall and transparently substitutes each image with a vision model's description.

上下文工程：按层级（规则文件→规格→源码→错误输出→对话历史）主动策划 agent 看到的信息，提升输出质量。受 addyosmani/agent-skills（88k★ MIT）启发。

Lightweight, route-preserving vision link for DSH: a configured vision model sees while the selected text model stays in control.