시각 도구 · DeepSeek Harness 플러그인 — 58 페이지

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

Local-first vision for text-only DeepSeek Harness agents: route image understanding to a local OpenAI-compatible vision model, return structured JSON evidence (summary / OCR / layout / semantics / visual / uncertainty).

Gives a text-only LLM vision capability

Per-model and per-route image-input (vision) checkboxes on the Models page's custom-provider cards, writing through the llm-pi-ai settings namespace.

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）。

DeepSeek Harness plugin: upload/paste images in the chat box; on send, transcribe via a vision model (dashscope) or offline Windows OCR, then inject the description into the message as 【解析了提供图片，图片内容是<描述>】.

Give text-only DeepSeek-V4-Pro real vision with zero new dependencies and DeepSeek-only routing: images are described by deepseek-v4-flash-vision-exp (your existing DEEPSEEK_API_KEY), then the text is handed to V4-Pro.

Give a text-only DeepSeek Harness model eyes: a proxy provider that delegates image understanding to any vision-capable model the harness can route, plus vision and vision_config tools.

DeepSeek Harness image recognition plugin: keep DeepSeek as the conversation brain while a configurable vision model (15+ providers, OpenAI-compatible + Anthropic) transcribes attached images. Configurable from Settings → Plugins.

常驻视觉服务：直连视觉模型（默认 opencode-go/minimax-m3，回退 zai-coding-cn/glm-4.6v）。describe_image / subagent_vision 工具 + 粘贴图片自动转译（llm/stream 钩子）+ 输入框视觉状态小胶囊与详情页（活动日志：指令/思考过程/输出）。零子代理、零 agent 上下文开销，按会话记忆窗支持视觉追问与验收。

Display companion for dsh-vision-router: keep Auto Vision routing, and put a square picture button on each original model row. Does nothing unless dsh-vision-router is installed.

让 DSH 里任何模型（包括 DeepSeek 这类纯文本模型）都能识别图片：识图技能 + 图片发送补丁，装上即用。

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.

Loss-aware, full-coverage image tiling for DeepSeek Harness vision models

DSH 模型视觉开关：设置页「模型视觉」行，为 llm-pi-ai 自定义路由下手写声明的模型切换 input 视觉模态（settings.mutate 官方通道，热生效）。

Windows-first vision suite for DeepSeek Harness.

Windows-first, non-invasive vision tools for DeepSeek Harness.

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.

Local-first visual acceptance and regression workbench for DeepSeek Harness

视觉层级构建参考

DeepSeek Harness (dsh) bundle plugin: vision_analyze tool lets text-only LLM agents read images via SenseNova VLM, with Schemastery config and single-source credentials

Voice for DeepSeek Harness — give text-only DeepSeek ears and a mouth: browser-native speech input (STT) + read-aloud (TTS), plus Whisper/TTS agent tools.

Voice for DeepSeek Harness backed by Xiaomi MiMo: browser-native 🎤/🔊 UI (MiMo TTS read-aloud) + voice_transcribe/voice_speak calling MiMo ASR/TTS directly, with a configurable voice map and in-conversation speech strips. Fork of zhuiyueya/dsh-voice (MIT).

Set a custom background wallpaper for the DeepSeek Harness desktop app — presets, image URL, upload, and an opacity slider