視覚ツール · DeepSeek Harness プラグイン — 10 ページ

Image understanding for text-only models via OpenRouter free vision models

Add no-key hosted or OpenAI-compatible vision to DeepSeek Harness while keeping the configured reasoning model

DSH Web GUI 皮肤插件：侧栏「皮肤」面板，支持自选本地图片 / URL 作为全屏背景，含预设、暗化、模糊、关闭皮肤，跨刷新持久化（外部 client 插件，dsh-web-ui 风格）。

Persistent workspace file tree for DSH Web: explorer + preview (text/docx/image/PDF/Markdown/CSV), multi-tab, inline editing, Git changes panel, file operations (rename/delete/copy/move/create), search filter, sorting

Pure React attachment atoms for the dsh web UI: draft-image rail, message image gallery, and original-image lightbox (zero cordis)

Dynamic attachment presentation plugin for conversation input, message-image, and trajectory image slots

Pure React attachment atoms for the dsh web UI: draft-image rail, message image gallery, and original-image lightbox (zero cordis)

Dynamic attachment presentation plugin for conversation input and message-image slots

Pure React attachment atoms for the dsh web UI: draft-image rail, message image gallery, and original-image lightbox (zero cordis)

Pure React attachment atoms for the dsh web UI: draft-image rail, message image gallery, and original-image lightbox (zero cordis)

Browser control: drives Edge/Chrome via the Chrome DevTools Protocol so the AI can automate the web (navigate, click, type, screenshot, evaluate JS). Settings live in the 'browser-control' namespace with a /api/browser status surface

Browser control: drives Edge/Chrome via the Chrome DevTools Protocol so the AI can automate the web (navigate, click, type, screenshot, evaluate JS). Settings live in the 'browser-control' namespace with a /api/browser status surface

Multimodal vision tool for dsh: analyze images or screen captures using DeepSeek vision models

DeepSeek Harness plugin that uses configured model providers for image analysis and context compaction.

DeepSeek Harness (DSH) vision bridge: let text-only models delegate image understanding to multimodal vision models via vision_describe tool, with persistent settings UI and primary/fallback auto-retry. 视觉桥：让纯文本模型通过 vision_describe 工具把图片理解委托给多模态视觉模型，带持久化设置界面与主备自动重试。

DeepSeek Harness 主题皮肤插件：给 DSH Web GUI 换背景（内置主题 / 静态图片 / 动态视频环绕跟随帧），设置页与侧边栏底部一键切换。

WeCom AI Bot text, image, and file channel bridge for DeepSeek Harness

Image viewer for the dock file explorer: renders PNG, JPEG, GIF, WebP, BMP, SVG, ICO and AVIF files.

DeepSeek Harness multi-model routing plugin: route tasks to custom specialist agents (vision / image generation / translation / speech / subagent) with per-agent providers & models, multimodal account sign-in, OAuth account pools, and realtime usage stats

DSH web plugin: a VSCode-like right sidebar (explorer / editor / terminal / git / browser), isolated per conversation session. Exposes a service for other plugins to register sidebar tabs and file viewers.

Model Context Protocol (MCP) Multi-Source Manager and Testing Workbench for DeepSeek Harness

Agent Skills and Rules Workbench for DeepSeek Harness better-sidebar

Generate images via API (image models filtered from those already set up in DSH) or local CLIs (MiniMax mmx), probe available backends, and display images inline in the chat.

DSH 对话体验增强套件 + 四工作台（零 DSH 源码改动，纯插件注入）。回合呈现：思考 / 工具调用聚合 chip（实时走秒 + 分步跟随滚动）+ 步骤卡 / 总结卡（回合收口才出现）+ 共享活动抽屉（思考语义分组 + 工具调用树）。正文增强：proto-tabs 可交互卡片 + diagram 流程图围栏（JSON 转 SVG）+ 生图画廊 + 重试行影子。界面与工具：会话头部视图标签上移右上角 + 桌面壳窗口控制留位与主题同步 + 对话截图（常驻无头浏览器渲染 markdown/shiki/mermaid，可内嵌本地 HTML）+ download 下载工具（wire 工具 + 实时进度条/速度/ETA）+ KR 对话双栏大盘（右栏保留任务 / 思考 / 工具调用 / 记忆四张卡，顶栏由 KR_PANEL_HEADER_VISIBLE 门控）。融合的原 dsh-triad 四工作台（座位与路由前缀原样保留，dsh-triad 自此退役）：自动沉淀的长期记忆、定时自动化、用量与账号趋势、技能与 MCP Server 管理。