image 65

Eyes for text-only DeepSeek Harness agents: built-in free vision chain + pixel-level tools. Remote vision providers, including the default free fallback, receive image content unless strict local-only use is configured.

Unified image understanding plugin for DeepSeek Harness (DSH). Visual twin adapter for native thumbnails + auto-analysis on any text-only model (incl. pi-ai providers); privacy/smart/strict routing; local tools (scan/OCR×4 engines (windows/macos/paddle/ra

DeepSeek brain + automatic image transcription for DeepSeek Harness: a deepseek-vision provider route that transcribes attached images to text before delegating to the text-only DeepSeek adapter. Official deepseek-v4-flash-vision-exp by default (a pure-te

Generate and display images inline in the DSH chat via API channels or local CLIs (mmx / codex / agy), and read images into structured JSON evidence (OCR / layout / semantics) on any model — with backend probing and a bundled recovery skill.

Eyes for text-only DeepSeek on DeepSeek Harness: image understanding via Doubao Web by default (zero cost, no API key), Antigravity IDE quota (flash/pro), Gemini API, or Cockpit proxy — model-invokable vision tool, wrapper adapters, evidence memory, and a

DeepSeek Harness 外置视觉模型插件：inspect_image 把本地图片或 http(s) 图片 URL 发给任意 OpenAI 兼容端点，视觉模型看图的文字回答直接带回对话；附 Web UI 设置栏。

Display images read by the read_image tool inside the DeepSeek Harness Web GUI: a DSH 0.2-compatible dedicated Read image row with a default-expanded message-style image card (PS-style transparency checkerboard) and an in-page full-resolution lightbox. 在

Auditable vision and cross-platform Computer Use runtime for DeepSeek Harness with source-preserving evidence.

DeepSeek vision bridge for the dsh web GUI: route image attachments to a vision model (pi-ai / llama.cpp Qwen3-VL) and continue the conversation with the text description on a text-only LLM (DeepSeek)

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

DeepSeek Harness plugin: let text-only models receive pasted images, and analyze them with a built-in OpenAI-compatible vision tool

Eyes for text-only DeepSeek on DeepSeek Harness: image understanding via Antigravity IDE quota (default, flash/pro), OpenAI-compatible VLM endpoints, Gemini API, or local Ollama — model-invokable vision tool, wrapper adapters for deepseek/opencode-go, evi

Persistent inline image, video, and audio previews for DeepSeek Harness Web conversations, with workspace-confined local reads and an optional ComfyUI proxy.

为 DeepSeek Harness 提供外部视觉模型能力：纯文本主模型通过 describe_image 工具调用外部视觉模型看图，拿到纯文字描述（多模态补全）。静态 Cordis 插件，随 DSH 启动自动加载。

Give DeepSeek Harness free eyes — natively. Paste or drop an image into the Web GUI composer: it appears as a thumbnail attachment (like a normal AI chat), and before the text-only DeepSeek model is called, the host automatically reads the image with the

Adaptive image routing for DeepSeek Harness: pass images directly to native multimodal models and transcribe them only for text-only models.

对话里的图值得被认真看：点一下全屏灯箱（缩放/拖拽平移/下载另存/左右翻页），一键把会话内的图收成可调列数的缩略图网格，或整块隐藏（不留空白）；本地图 TIFF/HEIC 自动无损转码，网图被防盗链拦截时自动转存本地显示。| Chat images get a proper look: full-screen lightbox (zoom / drag pan / download / prev-next), a conversation-wide thumbnail grid with adjustab

DeepSeek Harness plugin: a model-facing `vision` tool that describes and OCRs image files by calling the free Zhipu GLM vision API directly (no external CLI required).

Model-facing generate_image tool for DeepSeek Harness (dsh): generates brand-new images with the Google Gemini image-generation service via the Antigravity CLI, saves them, and returns the paths. Hot-pluggable — install with `dsh plugin --profile web add

聊天输入框图片工具(自研):上传图片(≤10MB 防烧 token) + 自定义区域截图(Electron desktopCapturer),注入 DSH 草稿图片轨。零外部依赖。

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。

Vision toolkit for DeepSeek Harness -- glance, ground, detect, crop CLI tools for text-only agents to understand images

DeepSeek Harness host plugin: serves a configured workspace/media folder over the web server under /media, so any conversation can display local images by referencing http://<host>:<port>/media/<relative-path>.

Region & sticker image skin for the DSH (DeepSeek Harness) web UI - paint the window, center, sidebar, welcome, right panel and composer with your own images, GIFs or videos, add two draggable stickers, and let the interface take its colour from your wall

Per-model vision (image input) toggle for DeepSeek Harness (dsh): list every configured model and flip a switch to enable/disable image support without hand-editing settings.yaml. 为 DeepSeek Harness 提供按模型的「支持图片」开关：无需手改 settings.yaml。

Bridges images into text for DeepSeek Harness: when the session's selected model cannot see images, a configured vision model describes them and the descriptions enter the durable session history as folded context rows — your message stays untouched.

DSH 静默视觉增强：主模型照常选择，图片自动交给固定视觉模型后以隐藏上下文返回主模型。

DeepSeek Harness (dsh) web plugin: paste images into the composer with thumbnail previews, route them as file attachments for text-only models, and show zoomable thumbnails in sent messages

DeepSeek Harness 视觉桥：自动发现你已配置的多模态模型，给纯文本主模型装上 vision 工具，识别结果以纯文本返回。零配置，一条命令安装。

dsh auto-vision bridge: paste an image into a text-only model's composer and a configured multimodal model transcribes it as text, automatically.

Image generation through the locally installed Doubao desktop app, driven over the Chrome DevTools Protocol — no API key, no headless browser.

DSH 生图插件 Ark9canvas：Agent 生图(默认需用户批准) + 浮窗生图工作台(比例/透明背景/批量/提示词库/多渠道/生成记录)，OpenAI 兼容图片 API 文生图/图生图。

DSH 插件：面向 AI 漫剧 / AI 微电影 / AI 视频创作的多项目无限画布编辑器 —— 画布节点编排、分门别类的素材库、全局素材搜索与跨项目软引用，配合 dsh-llm-multimodal 的生成工具与 media-preview 的聊天内预览。

Lets the agent put a picture into the conversation: a screenshot, the clipboard, or a file — every client renders it and the user can open it full screen.

DSH 视觉桥插件：自动区分多模态/文本模型。多模态模型直接看图；文本模型通过可配置的多模态端点（baseUrl + apiKey + model）代看，支持粘贴图片、read_image 工具、请求时图片转证据。

DSH plugin: read images through a user-configured vision-capable model when the main model route cannot accept image input

附件自动降级：DSH 图片准入（单边 ≤2000px / ≤3.5MB / ≤4000万像素）超限图片不再拒绝，用 sharp 自动缩小/重编码后入库，用户上传手机原图不再踩线。

Vision bundle + plugin for DeepSeek Harness: the describe_image tool reads local images and asks any configured multimodal route, with zero core changes

Renders chat Markdown image references () as a gallery under each closing assistant message

对话里的图值得被认真看：点一下全屏灯箱（缩放/拖拽平移/下载另存/左右翻页），一键把会话内的图收成可调列数的缩略图网格，或整块隐藏（不留空白）；本地图 TIFF/HEIC 自动无损转码，网图被防盗链拦截时自动转存本地显示。| Chat images get a proper look: full-screen lightbox (zoom / drag pan / download / prev-next), a conversation-wide thumbnail grid with adjustab

图片处理：读取图片尺寸/格式、缩放与格式转换（内置 ffmpeg，跨平台）

Give text-only models eyes: an analyze_image tool for DeepSeek Harness, backed by free Chinese vision APIs (GLM-4V-Flash / Qwen-VL) or any OpenAI-compatible vision endpoint. 给纯文本模型装上眼睛。

Request image optimization, intelligent retention, and observability for DeepSeek Harness (DSH)

DSH 图片插件，三个工具覆盖三种场景：ask_user_choice 图片/图文混合选择卡（Web GUI 渲染，可放大查看）+ show_images 回复内嵌图片（图文混排）+ save_received_images 盲模型收图存为工作区文件；聊天栏所有图片点击放大，支持滚轮缩放与拖拽平移。来源支持本地路径 / http(s) URL / base64 data URI。零 token 本地渲染，纯插件实现不改核心包。

DSH 图片/图文混合选项插件：ask_user_choice 工具让模型给用户提供带图片的选项（本地路径 / http(s) URL / base64 data URI），Web GUI 渲染图片选择卡，点击卡片可放大查看大图。纯插件实现，不改核心包。

DSH 插件：在聊天记录中自动将本地音视频/图片路径渲染为可播放的预览组件。When an assistant message or tool result contains a local media path, the path is replaced inline with a playable <audio>/<video>/<img> element backed by same-origin /api/media-preview/* route.

DSH profile plugin: converts read_image PNG/WebP attachments to opaque sRGB JPEG so LM Studio's openai-completions endpoint accepts them

DSH plugin: image understanding via any OpenAI-compatible vision API, paste-to-describe, and an animated whale-buddy desktop pet with status bubbles and a floating settings panel

Transparent image guard + vision analysis for DeepSeek Harness: text-only models read pasted images without the 400 session deadlock.

DeepSeek Harness plugin: local text-to-image with a canvas-style GUI, inpainting, DeepSeek vision review, OpenAI-compatible API option, and a Python Stable Diffusion worker (diffusers/MPS).