Vision tools · DeepSeek Harness plugins — Page 7

Use DeepSeek Harness directly in VS Code: secondary sidebar chat panel, @dsh built-in chat participant with @ file/session/agent mentions, multimodal image attachments, model/preset/permission/thinking controls, subagents, Cordis plugin approvals, and turn-level Git rollback (undo one turn, restore checkpoints).

为 deepseek-v4-flash-vision-exp 量身定制的大图智能识图插件：整图预检 → 本地 OCR+像素网格（文字）+ 兴趣点按比例切块（最长边 800）→ 汇总；保留全图 800x800 网格切块模式。不统计 token、不计算费用。v0.4.2：picturereader 适应性优化（共存分工引导/视觉端点复用/OCR venv 共享）。

DeepSeek Visionary native plugin for DeepSeek Harness: deepseek_vision / status / login / logout native tools plus the text-model image bridge, all backed by the visionary-server CLI (DeepSeek web vision model, no API key).

A desktop companion for DeepSeek Harness: its own persona, lorebook and multimodal API, a user profile distilled from the harness's own chat history, awareness of the running task, proactive chatter, reminders, voice, 8-24 frame sprite animations, screen reading and computer control behind a three-level permission, and a click-to-talk desktop window (WinForms sprite or Edge app window) next to the in-app floating pet.

Drop any file onto the DeepSeek Harness chat: non-image files are saved into the session workspace's .dsh-uploads/ folder and referenced from the composer as @.dsh-uploads/<name>, so the agent reads them with its file tools. Images keep going to the core's own attachment rail.

Extra media APIs for DeepSeek Harness: image generation, video generation, text-to-speech and transcription behind one settings panel, with adapters for OpenAI-compatible endpoints, Google Gemini/Veo, Replicate, fal, ElevenLabs, Fish Audio, MiniMax and any custom HTTP endpoint.

Skin Studio for DeepSeek Harness: a model-facing skin_studio tool that lets the agent design and one-click install launcher or frontend skins on user request — with a guided flow (ask requirements first; request an image from the user when the model cannot generate one, or call an image-generation tool when it can), local image inlining as data URIs, and instant apply through the DSH Launcher API.

DSH Web GUI right sidebar: VS Code-style file explorer plus source control (git status, stage/unstage/discard, commit, diff, commit graph, sync fetch/pull/push, branch/remote management, git init, multi-repo switch, color-coded file status, file-type icons, image preview) with an editable full-file diff view (hunk stage/revert), a syntax-highlighted editor (15 languages), file operations (new/move/copy/delete), a collapsible icon rail, a multi-tab bottom terminal (ConPTY), AI-generated commit messages from a host-configured model, and save.

图片理解桥接(dsh-vision-bridge 的中文界面分支):聊天里的图片交给你选定的视觉模型描述,非视觉模型也能"看图"。Fork of @goodandready/dsh-vision-bridge 0.4.5 (MIT) with a Chinese UI; the wire, routes and config file are unchanged.

A DeepSeek-style paperclip attach button in the DSH web composer; uploads files to the DSH host (model-aware: native image blocks for multimodal models, file paths for text-only models).

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

Drive OriginLab Origin scientific plotting from any AI via MCP - 62 tools: file import (CSV/TXT/XLSX, CJK-safe), plan-confirm flow with plan_hash staleness guard, domain templates (stacked spectra/XRD/dual-Y/forest/multi-panel), fine-grained editing (per-curve color/width/symbol, axis, legend anchors, page geometry, window management), editable OPJU delivery, deterministic readback verification, inline image preview, statistics batch (t/ANOVA/PCA/survival) with confidence notes, system diagnostics, scenario cookbook, version capability handshake, stable error codes with recovery map. Works standalone via npx (official MCP SDK) and inside DeepSeek Harness (DSH bundle).

Web GUI background plugin: per-area image groups with per-image display and rendering config (fill/fit/stretch/tile/center/custom + position, scale, size, repeat, rotation, radius, opacity, blur), slideshow playback, and local file/folder picking stored as IndexedDB blobs (no base64 packing); own settings section. TypeScript sources, esbuild-built bundle.

蓝色大肥鱼 · DeepSeek 峰谷提醒：北京时间「梁文峰/梁文谷」实时播报、切换前提前提醒、省 token 小贴士、自由拖动（纯前端，兼容 blue-fantasy 皮肤风格）

DeepSeek Harness web background settings plugin with solid color, image, and opacity controls.

DeepSeek 额度与用量仪表盘 —— DSH (DeepSeek Harness) Web 插件：余额还能撑几天、今日/本月花了多少、哪个模型和哪个会话最贵、图片/多模态用量、缓存省了多少、高峰挪到闲时能省多少

为 DeepSeek Harness 提供外挂识图模型：网页配置面板、发送图片识图自动回传、模型自主截图识图工具，支持多协议与一键安装。

Permission gateway for DSH (DeepSeek Harness): reviews tool calls per category (directory/command/read/image/edit/undo/subagent/doom loop) with global/project config, exceptions, quick-tool defaults, custom rules, a fallback policy for unmatched calls, bilingual approval modal and sandbox-upgrade flow. Config persisted at $DSH_HOME/dsh-permgate/config.json.

Image2 (gpt-image-2) generation plugin for DeepSeek Harness with simple relay settings and in-conversation image previews

Vision bridge plugin for DeepSeek Harness: when the main model has no vision, forward user images to a configurable OpenAI-compatible vision model and show results in a Web UI right panel. Host tool + browser half, distributed as a dsh bundle.

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

Replaces the official Models settings page to add per-model vision + reasoning-effort selectors for custom pi-ai providers

Free vision plugin for DeepSeek Harness (dsh): image understanding for text-only models with free-tier providers (Qwen3-VL-Flash / DeepSeek-OCR / Doubao). 免费视觉插件：纯文本模型看图能力，优先免费模型（通义千问 / 硅基流动 / 豆包）。

网页抓取与分析插件：抓取网页正文并给出统计与关键词分析，支持内容图片下载与识图分析器扩展接口。