276 results
DSH vision bridge (DSH >= 0.1.2-rc.1): when the selected chat model is text-only, attached images are described by a local Ollama VL model (qwen3-vl:8b) with keep_alive VRAM cooling. Install-time patch of dsh-api-session-controller prompt admission + runtime status companion.
DSH plugin: prompt the agent to dispatch image recognition to an opencode-go mimo-v2.5 subagent
DSH LLM adapter that transcribes native image attachments with Codex Luna before delegating to DeepSeek
DSH 视觉桥接插件：让无视觉能力的主模型看图（会话收图 + 自动转文字 + view_image 工具）
通过三方多模态 API 为纯文本 LLM 提供图像理解与文档理解能力（OCR、表格、公式、 版面解析、PDF/扫描件→Markdown、截图/照片/图表问答）。使用场景：用户粘贴或给出 图片/PDF 路径时解读内容、从文档提取文字与表格、对截图提问、发票/论文/票据解析。 触发词：图片理解、OCR、文档解析、截图内容、PDF 提取、表格识别、公式识别、票据识别、 扫描件、image、vision。实现为 Python 标准库 CLI（urllib/json/base64），零三方依赖。
DeepSeek Harness 宿主插件：让不支持图片输入的纯文本主模型也能"看图"（describe_image 工具 + 图片标记替换）
Unsloth for DeepSeek Harness - a pure-client tool plugin that connects the harness online model to a locally running Unsloth Desktop (Unsloth Studio) server for repetitive text and vision (OCR) labor.
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。
DeepSeek VisionPlus — official-grade vision extension for DeepSeek Harness. Routes image understanding to a free vision-model pool (Zhipu GLM, SiliconFlow Qwen) with automatic fallback, rate limiting, one-click platform tests and live status lines; text stays on DeepSeek. One-command install. MIT.
On-demand vision for text-only LLMs in DeepSeek Harness: keep pasted/attached images in the backend, and let the model call view_image to look at them anytime via any OpenAI-compatible vision endpoint — as if the model were natively multimodal.
Aura Vision — free vision OCR plugin for DeepSeek Harness web profile: Zhipu GLM-4V-Flash (free tier), adaptive tile recognition for long documents, history with favorites and Markdown/Excel/Word/PNG export.
Lightweight, route-preserving vision link for DSH: a configured vision model sees while the selected text model stays in control.
Let DeepSeek-V4-Pro (text-only) use V4-Flash-Vision-Exp for attached images. Mac/Windows/Linux.
DeepSeek Harness plugin: auto-route image-bearing requests to deepseek-v4-flash-vision-exp, then fall back to the original model.
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。
Model-facing analyze_image tool for the DeepSeek Harness: multi-modal image understanding via any OpenAI- or Anthropic-compatible vision API, with 8 analysis modes, local path / http(s) URL / data URL input, and a Web UI hint that guides image-incapable models to the reliable local-path route.
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 agent tools: slice a large image into labeled tiles (size or columns x rows) with an overview and a visual workbench; read_tiles selects only what the vision model needs.
DSH plugin: auto-tiles oversized chat images into labelled grid pieces (row/col metadata, overlap-aware, multi-image group isolation) so the DeepSeek vision model keeps fine detail instead of the ~800px downsample.
ClawShell vision re-implemented as DeepSeek Harness plugins: self-perception (sense), self-adaptation (adapt), swarm trust + ecological niche (swarm), plus tools.
让 DSH 直接操控你的浏览器，无需视觉能力：零 spawn 连接你已开启调试端口的 Chrome/Edge，截图、像素断言、执行页面 JS。Drive your already-running Chrome/Edge over CDP — screenshots, pixel assertions, page JS — no vision model, no per-use approval, zero child processes.
Vision for text-only DeepSeek Harness agents: delegate image reading to a one-shot subagent on a configurable vision route (MiniMax / Kimi / any OpenAI-compatible provider), keeping image bytes and the vision model's context out of the main session.
mm-vision (通感编码器) for DeepSeek Harness — give any text-only LLM the ability to see images via structured spatial text encoding. Registers the mm_vision tool.
dsh bundle: subagent_vision — delegate image reading to a vision-capable model from a text-only session, plus paste-to-path so pasted images reach the subagent as file paths.