插件目录 / page 172
全部插件 · 第 172 页
共 18502 个插件，每页 100 个。
@abdullaheltiby Silver ? Unverified Vision bridge for text-only models (DeepSeek, pi-ai routes): Gemini Flash describes attached images at the llm stream seam and provides a describe_image tool.
@lcb522 Silver ? Unverified DSH plugin: let text-only model sessions see images via a vision-model bridge (chat uploads + read_image)
@lynlicoder Silver ? Unverified 6 种检测方法融合决策（Tesseract OSD + 霍夫变换 + 四向试探 + 文字密度 + 区域 OSD + 多区域放大重检），自动检测并修正 PDF 页面方向（0°/90°/180°/270°）。100% 本地处理，零数据上传，生成可解释的 JSON 修正报告。
@jiawood2006 Silver ? Unverified Local OCR for PDFs/scans/images using macOS Vision. 本地OCR文档识别。
@doebkblcya Silver ? Unverified Dynamic vision bridge plugin for the DSH Web GUI: pasted images are recognized on demand via Bailian vision models (qwen3-vl-flash / qwen-vl-ocr)
@dreamrift Gold ? Unverified DeepSeek Harness 多模态视觉桥（dsh-vision-bridge）：贴图经 llm/stream 自动转 VL 文字描述（解决 UNSUPPORTED_CONTENT）+ view_image/ocr_image 主动视觉工具 + 原生多模态路由自动跳过（rc.7）；零依赖。Vision bridge for text-only DeepSeek models.
@xiyue718 Silver ? Unverified DSH 外接识图插件：当当前模型不支持识图时，调用外部视觉模型识别图片。 / External vision model plugin for DSH: recognize images when the active model lacks vision.
@tenglongbohai Gold ? Unverified DSH plugin: Add PaddleOCR vision capability to Agnes-2.5-Flash free model
@cyh12345678910 Silver ? Unverified Multi-backend vision plugin for DeepSeek Harness — API (OpenAI Vision) + CDP (Doubao bridge), cross-platform, cached, configurable
@avaritiachaos Silver ? Unverified Dynamic multimodal-to-text projection and cross-model vision bridge for DeepSeek Harness (dsh)
@uachargg Silver ? Unverified 
@huoer9527 Silver ? Unverified 
@harryli-7 Silver ? Unverified 
@soli0x4ea Silver ? Unverified 
@linktao159357 Silver ? Unverified 
@veryint Gold ? Unverified 
@seryta Gold ? Unverified 
@toryreina Silver ? Unverified 
@motongv Silver ? Unverified 给 DeepSeek Harness（DSH）加视觉能力的插件：装完就能拖图识别——图片交给视觉模型识别、文字交给 DeepSeek 推理，聊天入口完全不变。DSH 专属插件。
@thjyy Silver ? Unverified 
@zzdream67 Gold ? Unverified Let text-only models read images in DeepSeek Harness
@dddzzz123-dz Silver ? Unverified Image input fallback for DeepSeek Harness with native multimodal model detection and Volcengine Ark vision.
@co-elly Silver ? Unverified 👁️ Give your DeepSeek Harness the gift of sight — enables pure-text LLMs to analyze images via Zhipu's free GLM-4V-Flash vision model
@wangzhanchao883 Silver ? Unverified DeepSeek Harness plugin collection: self-developed DSH plugins (screenshot capture, OCR, Obsidian). ?? DSH ??????
@qedqcd Silver ? Unverified 
@mutton-hub Silver ? Unverified AdaTile-MCP: high-resolution image adaptive tiling MCP server for DeepSeek vision model (deepseek-v4-flash-vision-exp). L1-L6 pipeline (fastpath, saliency, adaptive tiling, Files API assembly, streaming VLM, rule-based merge) + eval harness. Setup: clone, run setup.bat, add your agent MCP config -> see README.
@ever0330 Silver ? Unverified Universal vision-model plugin for DeepSeek Harness, enabling image description using any configured vision model.
@carlmarkswx Silver ? Unverified Unified vision and image-generation plugin suite for DeepSeek Harness
@minglink Silver ? Unverified DeepSeek Harness 外挂视觉多模态与 OCR 桥接插件
@pipiwolve Gold ? Unverified Baidu cloud OCR bundle for DeepSeek Harness: drag images/PDFs in, OCR to Markdown with PaddleOCR-VL or Unlimited-OCR. 百度云 OCR 插件：拖入图片/PDF 识别为 Markdown 并写入本地文件。
@lasdrder0705 Gold ? Unverified dsh-std Community v0.15 ModelProvider: V4-Pro with Flash-Vision captions. Install adapter-dsh first.
@castem114 Silver ? Unverified 为 DeepSeek Harness（DSH）Web 量身打造的双插件扩展，为纯文本模型补齐"视觉"短板
@niushuanan Silver ? Unverified Let vision-capable models read native attachments while giving text-only models an image tool that supports follow-up questions.
@kl3jd Silver ? Unverified DSH plugin for local OCR: handwritten Chinese + math → Markdown with LaTeX. GPU / CPU backends.
@xby-skill Gold ? Unverified 输入一张图像，对其中的人头人体进行检测，输出图片中所有目标的检测框、置信度和标签。
@xby-skill Gold ? Unverified 识别银行卡号、发卡银行和卡类型，使用 Luhn 算法校验卡号有效性。
@avdergh Silver ? Unverified Preserve anti-aliased outlines while cutting chroma-backed game assets. CLI + MCP + Codex plugin.
@xby-skill Gold ? Unverified 对包含主体物体的图像进行标签识别，输出主体物体的类别标签，目前已经覆盖了5万多类的物体类别。
@xby-skill Gold ? Unverified 对含有动物的图像进行标签识别，无需任何额外输入，输出动物的类别标签。
@xby-skill Gold ? Unverified 兼顾速度与精度的文字识别。输入包含文本的图像，自动检测并识别内容。适用于各类文档、广告牌、屏幕截图等场景。
@xby-skill Gold ? Unverified 输入常见验证码图片，返回验证码文本内容。
@xby-skill Gold ? Unverified 检测各类通用场景中出现的火焰，最佳使用场景：安防摄像头、交通摄像头视角。
@xby-skill Gold ? Unverified 识别机动车行驶证的号牌号码、车辆类型、所有人、住址、品牌型号、发动机号码、车辆识别代号等信息，支持自动方向检测和主副页过滤。
@xby-skill Gold ? Unverified 输入一张图像，对其中的香烟目标进行检测，输出图片中所有目标的检测框、置信度和标签。
@xby-skill Gold ? Unverified 识别护照号码、中文姓名、英文姓名、性别、国籍、出生日期、签发日期、有效期至、签发地点等信息，支持MRZ机读码解析。
@xby-skill Gold ? Unverified 识别车牌号、车牌颜色、单/双层车牌、位置框。
@xby-skill Gold ? Unverified 识别营业执照的统一社会信用代码、名称、法定代表人、注册资本、成立日期、经营范围、登记机关和住所地址。
@xby-skill Gold ? Unverified 识别植物名称（或所属科, 属, 种或亚种）。
@xby-skill Gold ? Unverified 识别昆虫或其他节肢动物名称（或所属目, 科, 属, 种）。
@xby-skill Gold ? Unverified 识别身份证正面（姓名、性别、民族、出生日期、住址、身份证号）和背面（签发机关、有效期限），自动判断正反面并校验身份证号有效性。
@xby-skill Gold ? Unverified 识别驾驶证主页（证号、姓名、性别、国籍、住址、出生日期、准驾车型、初次领证日期、有效期限）和副页（档案编号）。
@iguanren Gold ? Unverified 让 DeepSeek Harness 纯文本模型也能识图：默认推荐智谱 GLM 免费视觉模型
@choco-zz Gold ? Unverified Drop historical images before DeepSeek Harness relay requests; keep native vision on the newest image.
@xiaxi626 Silver ? Unverified dsh-math-input 是一个 DeepSeek Harness 插件，提供零 token 消耗的离线数学输入能力。核心功能包括：手写笔迹识别（基于 ONNX 模型 + 束搜索）、LaTeX 自动修复与渲染（KaTeX）、以及手写画板 UI。所有推理在浏览器端完成，不依赖远程 API。
@wjling Silver ? Unverified dsh视觉助手，辅助没有多模态的主模型识别图片内容
@ryasrk Gold ? Unverified Patch custom provider vision model
@shainedemo Gold ? Unverified Give text-only DeepSeek-V4-Pro vision — zero new dependencies, DeepSeek-only routing (images described by deepseek-v4-flash-vision-exp, then answered by V4-Pro).
@shenhuanageshei Gold ? Unverified DeepSeek Harness plugin - session screenshots route by model capability: inline for vision models, VLM-read for text-only models (tool/auto modes)
@jonah791 Silver ? Unverified 多模态视觉插件（辅助通道）：把本地图片喂给 OpenAI 兼容 VLM（默认 qwen3.8-flash）做读图/描述/双图对比——批量审图/省主会话上下文；主会话亲眼看图走官方 read_image（原生多模态）
@alaxrpg Gold ? Unverified DSH 插件：为纯文本模型提供视觉能力，支持任意 OpenAI 兼容多模态 API
@mengruoa Silver ? Unverified 
@wuyiwuaa1-ship-it Gold ? Unverified DSH plugin that converts read_image PNG/WebP attachments to JPEG for LM Studio compatibility.
@mengruoa Silver ? Unverified DSH plugin: auto-install rembg and expose a background-removal tool to the LLM
@123twtd Gold ? Unverified DeepSeek Harness (dsh) 模型视觉开关插件：为自定义模型声明图像输入能力（写入 llm-pi-ai 配置），适配 0.1.x
@xby-skill Gold ? Unverified 对图像进行ImageNet 1000类分类，返回Top-5类别和置信度。
@xby-skill Gold ? Unverified 高精度文字识别。输入包含文本的图像，自动检测并识别内容。适用于各类文档、广告牌、屏幕截图等场景。
@xby-skill Gold ? Unverified 输入包含手写文本的图像，自动检测文本行并识别内容。适用于手写笔记、签名、手写表单等。
@xby-skill Gold ? Unverified 输入一张图像，对其中的人体、头部和安全帽进行检测，输出图片中所有目标的检测框、置信度和标签。
@xby-skill Gold ? Unverified 输入一张图像，对其中的手机进行检测，输出图片中所有目标的检测框、置信度和标签。
@xby-skill Gold ? Unverified 输入一张图像，对其中的人/宠物/车/火焰/纸箱进行检测，输出图片中所有目标的检测框、置信度和标签。
@cbg33695 Gold ? Unverified Let a text-only model see the screen: capture desktop/window/region, vision transcription, a few minutes of rolling screen memory, exact local pixel diff, and vision self-calibration. Windows only. Experimental - the README states what is measured and what is not.
@kim1232aa Gold ? Unverified DeepSeek Harness plugin: MxPage ecommerce product-image pipeline (analyze → plan → VPA → generate).
@sky1ine139 Gold ? Unverified DSH 截图提问插件：原生框选截图直接进输入框，全程不落盘（DeepSeek Harness plugin, Windows）
@mafeis Gold ? Unverified DSH plugin: trims historical images to the provider image budget before sending, learns the image cap from HTTP 400 responses, and retries with fewer images. Reduces vision-token usage and prevents image-heavy sessions from failing.
@zhiwuli0228 Gold ? Unverified DeepSeek Harness plugin: digest prompt images through a vision model and expose a describe_image tool — without ever changing the session's model.
@m0r1c Gold ? Unverified Fixes the error “This turn failed400: 'url' field must be a base64 encoded image.” when using local vision‑enabled models from LM Studio.
@pureexe Gold ? Unverified DeepSeek Harness LLM adapter that caps images per request to 1, for self-hosted vLLM backends that reject prompts with more than one image
@danhcng3822f Gold ? Unverified DeepSeek Harness plugin providing Add photos, Add files, and Uploaded files in chat UI with vision validation and session isolation
@davidekingsss Gold ? Unverified Autonomous screen vision for DeepSeek Harness on macOS and Windows: the agent captures the screen and receives the image itself in one call.
@lonelyherbivore Silver ? Unverified DSH Web plugin: ask the user a question with an image, annotate in the browser, get back structured annotations in source-image pixel coordinates.
@noahchalifour Gold ? Unverified DeepSeek Harness plugin: downscale images on Claude requests so many-image requests never hit the 2000px limit
@vitas Gold ? Unverified Local OCR for DeepSeek Harness: Apple Vision on macOS, no install, no network, no capability claims
@yuehancn Gold ? Unverified Extract text from images and PDFs in DeepSeek Harness: read a born-digital text layer, or OCR a scan with any engine you configure.
@lemonhall Silver ? Unverified 抠人物的方法与实测边界：模型代际、rembg 的 1024 陷阱、官方 2048 HR-matting 的软 alpha、8GB 显卡跑不了 2048 的原因，以及用「半透明像素占比」客观量发丝。附两个通用脚本（GPU 快路线 / 官方高清路线）。DSH 与 Codex 都能用。
@bauerelizabeth07139 Silver ? Unverified DeepSeek Harness plugin and OpenCode MCP server: describe an image with StepFun step-3.7-flash, so a text-only model can see. Exposes mcp__opencode_eyes__describe_image.
@bauerelizabeth07139 Silver ? Unverified DeepSeek Harness plugin and OpenCode MCP server: describe an image with any of seven NVIDIA NIM vision models, rotating several API keys. Exposes mcp__opencode_eyes_nvidia__describe_image.
@bauerelizabeth07139 Silver ? Unverified DeepSeek Harness plugin and OpenCode MCP server: local PP-OCRv6 Medium OCR (PaddleOCR on ONNX Runtime, CPU) for images and multi-page PDFs. Exposes mcp__ppocr__ocr_image and mcp__ppocr__ocr_pdf.
@senyayume Silver ? Unverified DSH plugin: inline each image block only in the request where it first appears, then swap it for the harness's own placeholder text
@ssrs-pro-max Gold ? Unverified DeepSeek Harness plugin: shows oversized images to the model at readable resolution by tiling them before admission.
@lilinhua0413 Gold ? Unverified Relabels DSH's normalized WebP request images as image/png so local OpenAI-compatible vision servers (LM Studio, llama.cpp) accept them.
@longyu065 Silver ? Verified No description provided.
@dmsobtl Silver ? Verified DSH plugin: multi-market financial data — A-shares, US stocks, crypto quotes/K-lines/stock screening/fundamentals.
@WODE25500 Silver ? Verified DSH SQLite data tool: list databases, tables, schema and run read-only SQL over local .db files. Zero-dependency (node:sqlite). DeepSeek Harness plugin.
@TikaFlow ? Verified 为所有非官方（自定义）提供商的模型自动填充推理级别，数据来自 models.dev。 / Automatically fills in reasoning-effort levels for models from unofficial (custom) providers, with data sourced from models.dev.
@rjn32s Gold ? Verified RDAP-backed whois tool plugin for DeepSeek Harness (dsh) — look up domain registration data as a model tool
@guannan1031 Silver ? Verified 电商经营驾驶舱 · Ecommerce Business Cockpit — DeepSeek Harness 演示插件。演示界面与数据校验层开源；客户数据映射、指标口径与诊断服务私有。
@satan9394 ? Verified DSH skill: SQL 优化, EXPLAIN 分析与索引策略（受 wshobson/agents 38k★ 启发）
@satan9394 ? Verified DSH skill: 数据库表设计，主键/规范化/索引/数据类型/约束/性能模式，PostgreSQL 重点（受 wshobson/agents 38k★ 启发）
@uckkk Gold ? Verified SQL 格式化：子句换行、关键字大写
@uckkk Gold ? Verified JSON 转 SQL