{ "$schema": "https://deeplugin.store/data/launch-packet.schema.json", "version": 1, "plugin": { "id": "deeplugin-8586b3b02faa16fda2b0", "name": "picturereader", "author": "jing-hy", "category": "tools", "version": null, "homepage": "https://github.com/jing-hy/picturereader", "firstObservedAt": "2026-08-16T09:56:16Z", "lastChangedAt": "2026-08-16T09:56:16Z" }, "links": { "detail": "https://deeplugin.store/plugins/deeplugin-8586b3b02faa16fda2b0.html", "feed": "https://deeplugin.store/plugins/deeplugin-8586b3b02faa16fda2b0.atom.xml", "newFeed": "https://deeplugin.store/feeds/new.atom.xml", "updatedFeed": "https://deeplugin.store/feeds/updated.atom.xml" }, "install": { "spec": "github:jing-hy/picturereader", "command": "dsh plugin --profile web add github:jing-hy/picturereader", "agentRequest": "请先在 deeplugin.store 中检索 Registry ID deeplugin-8586b3b02faa16fda2b0,展示来源和精确安装 spec,为 web profile 生成安装计划;等我明确批准后再安装。" }, "badge": { "markdown": "[![Listed on deeplugin.store](https://deeplugin.store/assets/deeplugin-listed.svg)](https://deeplugin.store/plugins/deeplugin-8586b3b02faa16fda2b0.html)", "html": "\"Listed" }, "posts": { "zh": "picturereader 已收录到 DeepSeek Harness Plugin Store。给纯文本模型的\"读图\"能力:图片降分辨率+降色深+结构/色彩指纹渲染成文本网格喂回对话,模型像多模态一样自主缩放、取样、OCR 读图;纯本地零外部模型依赖,附读图方法论 skill 与可选 PaddleOCR。\n\n查看来源与安装:https://deeplugin.store/plugins/deeplugin-8586b3b02faa16fda2b0.html", "en": "picturereader is now listed on the DeepSeek Harness Plugin Store. Image \"reading\" for text-only models: downscale + reduce color depth + structure/color fingerprints into text grids fed back to the conversation, letting the model zoom, sample and OCR autonomously like a multimodal model; fully local with zero external model dependency, ships an image-reading methodology skill and optional PaddleOCR.\n\nSource and install details: https://deeplugin.store/plugins/deeplugin-8586b3b02faa16fda2b0.html" } }