[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-lifan0127--ai-research-assistant":3,"tool-lifan0127--ai-research-assistant":61},[4,18,26,36,44,52],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",147882,2,"2026-04-09T11:32:47",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",108111,"2026-04-08T11:23:26",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":10,"last_commit_at":50,"category_tags":51,"status":17},4487,"LLMs-from-scratch","rasbt\u002FLLMs-from-scratch","LLMs-from-scratch 是一个基于 PyTorch 的开源教育项目，旨在引导用户从零开始一步步构建一个类似 ChatGPT 的大型语言模型（LLM）。它不仅是同名技术著作的官方代码库，更提供了一套完整的实践方案，涵盖模型开发、预训练及微调的全过程。\n\n该项目主要解决了大模型领域“黑盒化”的学习痛点。许多开发者虽能调用现成模型，却难以深入理解其内部架构与训练机制。通过亲手编写每一行核心代码，用户能够透彻掌握 Transformer 架构、注意力机制等关键原理，从而真正理解大模型是如何“思考”的。此外，项目还包含了加载大型预训练权重进行微调的代码，帮助用户将理论知识延伸至实际应用。\n\nLLMs-from-scratch 特别适合希望深入底层原理的 AI 开发者、研究人员以及计算机专业的学生。对于不满足于仅使用 API，而是渴望探究模型构建细节的技术人员而言，这是极佳的学习资源。其独特的技术亮点在于“循序渐进”的教学设计：将复杂的系统工程拆解为清晰的步骤，配合详细的图表与示例，让构建一个虽小但功能完备的大模型变得触手可及。无论你是想夯实理论基础，还是为未来研发更大规模的模型做准备",90106,"2026-04-06T11:19:32",[35,15,13,14],{"id":53,"name":54,"github_repo":55,"description_zh":56,"stars":57,"difficulty_score":10,"last_commit_at":58,"category_tags":59,"status":17},4292,"Deep-Live-Cam","hacksider\u002FDeep-Live-Cam","Deep-Live-Cam 是一款专注于实时换脸与视频生成的开源工具，用户仅需一张静态照片，即可通过“一键操作”实现摄像头画面的即时变脸或制作深度伪造视频。它有效解决了传统换脸技术流程繁琐、对硬件配置要求极高以及难以实时预览的痛点，让高质量的数字内容创作变得触手可及。\n\n这款工具不仅适合开发者和技术研究人员探索算法边界，更因其极简的操作逻辑（仅需三步：选脸、选摄像头、启动），广泛适用于普通用户、内容创作者、设计师及直播主播。无论是为了动画角色定制、服装展示模特替换，还是制作趣味短视频和直播互动，Deep-Live-Cam 都能提供流畅的支持。\n\n其核心技术亮点在于强大的实时处理能力，支持口型遮罩（Mouth Mask）以保留使用者原始的嘴部动作，确保表情自然精准；同时具备“人脸映射”功能，可同时对画面中的多个主体应用不同面孔。此外，项目内置了严格的内容安全过滤机制，自动拦截涉及裸露、暴力等不当素材，并倡导用户在获得授权及明确标注的前提下合规使用，体现了技术发展与伦理责任的平衡。",88924,"2026-04-06T03:28:53",[14,15,13,60],"视频",{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":66,"readme_en":67,"readme_zh":68,"quickstart_zh":69,"use_case_zh":70,"hero_image_url":71,"owner_login":72,"owner_name":73,"owner_avatar_url":74,"owner_bio":75,"owner_company":76,"owner_location":77,"owner_email":78,"owner_twitter":79,"owner_website":80,"owner_url":81,"languages":82,"stars":105,"forks":106,"last_commit_at":107,"license":108,"difficulty_score":32,"env_os":109,"env_gpu":110,"env_ram":110,"env_deps":111,"category_tags":117,"github_topics":119,"view_count":32,"oss_zip_url":78,"oss_zip_packed_at":78,"status":17,"created_at":125,"updated_at":126,"faqs":127,"releases":128},5882,"lifan0127\u002Fai-research-assistant","ai-research-assistant","Aria is Your AI Research Assistant Powered by GPT Large Language Models","Aria 是一款专为文献管理工具 Zotero 打造的 AI 研究助手插件，旨在利用大语言模型（如 GPT-4）提升学术科研效率。它主要解决了研究人员在阅读海量文献时面临的整理困难、笔记繁琐及深度分析耗时等痛点，让用户能直接在熟悉的文献库环境中与 AI 进行自然对话。\n\n这款工具特别适合高校学者、研究生及各类需要处理大量学术资料的研究人员使用。Aria 的核心亮点在于其与 Zotero 的深度集成：用户只需拖拽文献条目或集合即可快速引用上下文；支持作者、标签等内容的智能自动补全；更独特的是，它集成了 GPT-4 Vision 视觉分析能力，允许用户在 PDF 中框选图表或特定区域，让 AI 直接解读视觉信息。此外，所有的对话内容均可一键保存为笔记或注释，无缝融入现有的知识管理体系。通过简单的配置和快捷键操作，Aria 将强大的生成式 AI 能力转化为触手可及的科研助力，帮助用户更专注于核心创新工作。","# A.R.I.A. (Aria) - Your AI Research Assistant\n\n[![License](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Flifan0127\u002Fai-research-assistant)](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fblob\u002Fmaster\u002FLICENSE)\n[![Using Zotero Plugin Template](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FUsing-Zotero%20Plugin%20Template-blue?style=flat-square&logo=github)](https:\u002F\u002Fgithub.com\u002Fwindingwind\u002Fzotero-plugin-template)\n[![Latest release](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Flifan0127\u002Fai-research-assistant)](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases)\n![Downloads latest release](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fdownloads\u002Flifan0127\u002Fai-research-assistant\u002Flatest\u002Ftotal)\n\nAria is a Zotero plugin powered by Large Language Models (LLMs). A-R-I-A is the acronym of \"AI Research Assistant\" in reverse order.\n\nPlease make sure to choose the correct version based on your Zotero version:\n\n- Zotero 6: https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.8.0\n- Zotero 7: https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Flatest\n\n![Aria](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_c9c26f10845e.png)\n\n## Quickstart\n\nThe easist way to get started with Aria is to try one of the interactive prompts in the prompt library.\n\n![Prompt Library](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_42e07cc45787.gif)\n\n## Features\n\n### Use Drag-and-Drop to Reference Your Zotero Items and Collections\n\n![Drag and Drop](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_0861505b00cd.gif)\n\n### Autocompletion for Creators (Authors), Tags, Items and More\n\n![Autocompletion](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_94ade1795112.gif)\n\n### Visual Analysis (GPT-4 Vision)\n\n[_How to use Zotero area annotation to create a draggable area in PDF?_](https:\u002F\u002Fyoutu.be\u002FlGeJCsNHBR4?si=sCnnaZCYI2CN8NJ3&t=132)\n\n![Visual Analysis](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_396da17902b1.gif)\n\n### Save Chat as Notes and Annotations\n\n![Save Chat as Notes and Annotations](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_00b6d7e2d7ae.gif)\n\n\n## Zotero and GPT Requirements\n\n- Please note separate releases are available for Zotero 6 and 7:\n  - Zotero 6: https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.8.0\n  - Zotero 7: https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002Fv0.7.4\n- Aria requires the OpenAI GPT-4 model family. ([how can I access GPT-4?](https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F7102672-how-can-i-access-gpt-4))\n- The visual analysis feature requires the preview access to the GPT-4 Vision model.\n\n## Installation\n\nFor a detailed walkthrough of the installation process, please check out: https:\u002F\u002Ftwitter.com\u002FMushtaqBilalPhD\u002Fstatus\u002F1735221900584865904 (credit: __Mushtaq Bilal, PhD__ - Syddansk Universitet)\n\n- Download the latest release (.xpi file) from GitHub: https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Flatest\n- In Zotero select \u003Cem>Tools\u003C\u002Fem> from the top menu bar, and then click on \u003Cem>Addons\u003C\u002Fem>.\n- On the Add-ons Manager panel, click the gear icon at the top right corner and select \u003Cem>Install Add-on From File\u003C\u002Fem>\n- Select the .xpi file you just downloaded and click \u003Cem>Open\u003C\u002Fem> which will start the installation process.\n\n## Quickstart\n\nBy default, Aria can be activated by clicking the ![Aria launch button](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_4f61f5202ef6.png) button on Zoterol toolbar or through the \"Shift + R\" shortcut.\n\nBefore using Aria, you need to provide an [OpenAI API Key](https:\u002F\u002Fplatform.openai.com\u002Faccount\u002Fapi-keys). Follow the in-app instruction to add a key and \u003Cb>restart Zotero\u003C\u002Fb>. ([screenshots](docs\u002Fconfiguration.md))\n\nAfter restart, you should see the activated Aria window (as shown above) and can start using it through conversations.\n\n\n## Preferences\n\nAria is configurable through Edit > Preferences > Aria. Please note that some changes require Zotero restart.\n\n- __Model Selection__: Choose between the base GPT-4 model and the new GPT-4 Turbo model (Preview).\n- __Zoom Level__: Adjust the zoom level to fit your screen resolution \n- __Keyboard shortcut__: Change the keyboard shortcut combination to better fit your workflow.\n\n![Aria](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_ae7e15a395cd.png)\n\n## Update\n\n- Aria can perform automatic update when internet access is available. To check for available update, select \u003Cem>Tools\u003C\u002Fem> from the top menu bar, and then click on \u003Cem>Addons\u003C\u002Fem>.\n- To manually update ARIA, click \u003Cem>More\u003C\u002Fem> under Aria and then click the gear icon at the top right corner. Select \u003Cem>Check for Updates\u003C\u002Fem>. ([screenshots](docs\u002Fupdate.md))\n\n## Limitations\n\nThe following are known limitations based on user feedback.\n\n- Currently Aria can query your Zotero library through the Zotero search API. The ability to query the Zotero SQLite database for document count and other metrics will be delivered in a future release.\n- Aria has limited awareness of your Zotero application state (selected item, current tab, highlighted text). However, you can use the drag-n-drop and the autocompeltion features to provide such context within your message.\n\n## Troubleshooting\n\n  Interaction with Zotero, in an open conversational manner and through a probabilistic model, can lead to many different, often unexpected outcomes. If you experience any error, please create an GitHub issue with a screenshot of the error message from your Aria chat window. Thank you!\n\n- \"Agent stopped due to max iterations\": For certain questions, the bot will make multiple API calls iteratively for response synthesis. Sometimes, it may fail to produce an answer before reaching the max iterations.\n\n- Aria tab not in Preferences panel: You may choose the __Advanced__ tab in Preferences and open the __Configuration Editor__ Under Advanced Configuration. From there, please search for \"aria\" and then double-click on the \"extensions.zotero.aria.OPENAI_API_KEY\" entry to add your OpenAI API Key.\n\n## Development\n\nRefer to the [Zotero Plugin Development](https:\u002F\u002Fwww.zotero.org\u002Fsupport\u002Fdev\u002Fclient_coding\u002Fplugin_development) guide to find instructions on how to setup the plugin in your local environment.\n\n## Feedback\n\nYou can now submit feedback and share your chat session to help improve Aria. Let's make Aria better together!\n\n![Visual Analysis](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_e10f6c017632.gif)\n\n\n","# A.R.I.A.（咏叹调）—— 你的 AI 研究助手\n\n[![许可证](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Flifan0127\u002Fai-research-assistant)](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fblob\u002Fmaster\u002FLICENSE)\n[![使用 Zotero 插件模板](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FUsing-Zotero%20Plugin%20Template-blue?style=flat-square&logo=github)](https:\u002F\u002Fgithub.com\u002Fwindingwind\u002Fzotero-plugin-template)\n[![最新版本](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Flifan0127\u002Fai-research-assistant)](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases)\n![最新版本下载量](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fdownloads\u002Flifan0127\u002Fai-research-assistant\u002Flatest\u002Ftotal)\n\nAria 是一款由大型语言模型（LLMs）驱动的 Zotero 插件。A-R-I-A 是“AI Research Assistant”（AI 研究助手）的反向缩写。\n\n请根据您的 Zotero 版本选择正确的版本：\n\n- Zotero 6：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.8.0\n- Zotero 7：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Flatest\n\n![Aria](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_c9c26f10845e.png)\n\n## 快速入门\n\n开始使用 Aria 的最简单方式，就是尝试提示库中的交互式提示。\n\n![提示库](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_42e07cc45787.gif)\n\n## 功能\n\n### 拖放引用 Zotero 中的条目和收藏集\n\n![拖放](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_0861505b00cd.gif)\n\n### 创作者（作者）、标签、条目等的自动补全\n\n![自动补全](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_94ade1795112.gif)\n\n### 视觉分析（GPT-4 Vision）\n\n[_如何使用 Zotero 区域标注在 PDF 中创建可拖动区域？_](https:\u002F\u002Fyoutu.be\u002FlGeJCsNHBR4?si=sCnnaZCYI2CN8NJ3&t=132)\n\n![视觉分析](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_396da17902b1.gif)\n\n### 将聊天保存为笔记和标注\n\n![保存聊天为笔记和标注](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_00b6d7e2d7ae.gif)\n\n\n## Zotero 和 GPT 要求\n\n- 请注意，Zotero 6 和 7 分别有不同的发布版本：\n  - Zotero 6：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.8.0\n  - Zotero 7：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002Fv0.7.4\n- Aria 需要 OpenAI 的 GPT-4 模型系列。（[我如何访问 GPT-4？](https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F7102672-how-can-i-access-gpt-4)）\n- 视觉分析功能需要 GPT-4 Vision 模型的预览访问权限。\n\n## 安装\n\n有关安装过程的详细说明，请参阅：https:\u002F\u002Ftwitter.com\u002FMushtaqBilalPhD\u002Fstatus\u002F1735221900584865904（鸣谢：__穆什塔克·比拉尔博士__ - 南丹麦大学）\n\n- 从 GitHub 下载最新版本的 .xpi 文件：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Flatest\n- 在 Zotero 中，从顶部菜单栏选择\u003Cem>工具\u003C\u002Fem>,然后点击\u003Cem>附加组件\u003C\u002Fem>。\n- 在附加组件管理器面板中，单击右上角的齿轮图标，选择\u003Cem>从文件安装附加组件\u003C\u002Fem>。\n- 选择刚刚下载的 .xpi 文件并点击\u003Cem>打开\u003C\u002Fem>,安装过程将开始。\n\n## 快速入门\n\n默认情况下，可以通过点击 Zotero 工具栏上的 ![Aria 启动按钮](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_4f61f5202ef6.png) 按钮或使用“Shift + R”快捷键来激活 Aria。\n\n在使用 Aria 之前，您需要提供一个 [OpenAI API 密钥](https:\u002F\u002Fplatform.openai.com\u002Faccount\u002Fapi-keys)。按照应用内指示添加密钥，并\u003Cb>重启 Zotero\u003C\u002Fb>。（[截图](docs\u002Fconfiguration.md)）\n\n重启后，您应该会看到已激活的 Aria 窗口（如上所示），并可通过对话开始使用它。\n\n\n## 偏好设置\n\nAria 可通过编辑 > 偏好设置 > Aria 进行配置。请注意，某些更改需要重启 Zotero。\n\n- __模型选择__：在基础 GPT-4 模型和新的 GPT-4 Turbo 模型（预览版）之间进行选择。\n- __缩放级别__：调整缩放级别以适应您的屏幕分辨率。\n- __键盘快捷键__：更改键盘快捷键组合，使其更符合您的工作流程。\n\n![Aria](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_ae7e15a395cd.png)\n\n## 更新\n\n- 在有互联网连接的情况下，Aria 可以自动更新。要检查是否有可用更新，从顶部菜单栏选择\u003Cem>工具\u003C\u002Fem>,然后点击\u003Cem>附加组件\u003C\u002Fem>。\n- 若要手动更新 ARIA，点击 Aria 下方的\u003Cem>更多\u003C\u002Fem>,然后单击右上角的齿轮图标。选择\u003Cem>检查更新\u003C\u002Fem>。（[截图](docs\u002Fupdate.md)）\n\n## 限制\n\n以下是根据用户反馈已知的限制。\n\n- 目前，Aria 可以通过 Zotero 搜索 API 查询您的 Zotero 库。未来版本将支持查询 Zotero SQLite 数据库以获取文档数量和其他指标。\n- Aria 对您的 Zotero 应用程序状态（选中项、当前标签页、高亮文本）的感知有限。不过，您可以使用拖放和自动补全功能，在消息中提供这些上下文信息。\n\n## 故障排除\n\n与 Zotero 的开放式对话以及基于概率模型的交互，可能会导致许多不同且常常意想不到的结果。如果您遇到任何错误，请在 GitHub 上提交一个问题，并附上 Aria 聊天窗口中错误消息的截图。谢谢！\n\n- “代理因达到最大迭代次数而停止”：对于某些问题，机器人会多次迭代调用 API 来合成回复。有时，在达到最大迭代次数之前，它可能无法生成答案。\n- Aria 选项卡未出现在偏好设置面板中：您可以在偏好设置中选择__高级__选项卡，并打开__配置编辑器__。在高级配置中，搜索“aria”，然后双击“extensions.zotero.aria.OPENAI_API_KEY”条目，以添加您的 OpenAI API 密钥。\n\n## 开发\n\n请参考 [Zotero 插件开发](https:\u002F\u002Fwww.zotero.org\u002Fsupport\u002Fdev\u002Fclient_coding\u002Fplugin_development)指南，了解如何在本地环境中设置插件。\n\n## 反馈\n\n现在您可以提交反馈并分享您的聊天记录，以帮助改进 Aria。让我们一起让 Aria 更加出色吧！\n\n![视觉分析](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_readme_e10f6c017632.gif)","# A.R.I.A. (Aria) 快速上手指南\n\nAria 是一款基于大语言模型（LLM）的 Zotero 插件，旨在辅助学术研究。它支持拖拽引用文献、自动补全、视觉分析（GPT-4 Vision）以及将对话保存为笔记等功能。\n\n## 环境准备\n\n在开始之前，请确保满足以下系统和账号要求：\n\n*   **Zotero 版本**：\n    *   **Zotero 6**：需下载 [0.8.0 版本](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.8.0)\n    *   **Zotero 7**：需下载 [最新版本](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Flatest)\n*   **OpenAI 账号**：\n    *   必须拥有 **GPT-4** 系列模型的访问权限。\n    *   若使用“视觉分析”功能，需具备 **GPT-4 Vision** 的预览访问权限。\n    *   准备好您的 [OpenAI API Key](https:\u002F\u002Fplatform.openai.com\u002Faccount\u002Fapi-keys)。\n\n## 安装步骤\n\n1.  **下载插件**：\n    访问 GitHub Releases 页面，根据您的 Zotero 版本下载对应的 `.xpi` 文件。\n\n2.  **打开插件管理器**：\n    启动 Zotero，点击顶部菜单栏的 `Tools`，然后选择 `Addons`。\n\n3.  **安装文件**：\n    在插件管理器面板右上角点击齿轮图标 ⚙️，选择 `Install Add-on From File...`。\n\n4.  **确认安装**：\n    在文件选择对话框中找到并选中刚才下载的 `.xpi` 文件，点击 `Open`。按照提示完成安装。\n\n5.  **配置 API Key**：\n    *   安装完成后，根据应用内指引输入您的 OpenAI API Key。\n    *   **重要**：配置完成后必须**重启 Zotero** 才能生效。\n    *   *备用方案*：如果偏好设置中未显示 Aria 选项，可进入 `Edit` > `Preferences` > `Advanced` > `Config Editor`，搜索 `aria`，双击 `extensions.zotero.aria.OPENAI_API_KEY` 手动填入 Key。\n\n## 基本使用\n\n安装并重启 Zotero 后，即可通过以下方式启动和使用 Aria：\n\n1.  **启动 Aria**：\n    *   点击 Zotero 工具栏上的 **Aria 图标**。\n    *   或使用默认快捷键 `Shift + R`。\n\n2.  **开始对话**：\n    *   启动后会出现 Aria 对话窗口。\n    *   **推荐新手操作**：点击提示词库（Prompt Library）中的交互式提示，快速体验功能。\n\n3.  **核心功能示例**：\n    *   **引用文献**：直接将 Zotero 中的条目或合集拖拽（Drag-and-Drop）到对话窗口中作为上下文。\n    *   **自动补全**：输入作者、标签或条目名称时，插件会自动提供补全建议。\n    *   **视觉分析**：利用 GPT-4 Vision，结合 Zotero 的 PDF 区域标注功能，对图表或特定文本区域进行可视化分析。\n    *   **保存结果**：对话内容可直接保存为 Zotero 笔记（Notes）或 PDF 注释（Annotations）。\n\n4.  **个性化设置**：\n    进入 `Edit` > `Preferences` > `Aria`，您可以：\n    *   切换模型（如 GPT-4 或 GPT-4 Turbo）。\n    *   调整窗口缩放比例。\n    *   修改启动快捷键。","博士生李明正在撰写关于“大语言模型在医疗诊断中应用”的综述论文，需要快速梳理 Zotero 库中上百篇 PDF 文献的核心观点。\n\n### 没有 ai-research-assistant 时\n- **人工阅读效率低**：必须逐篇打开 PDF 手动通读摘要和结论，耗时数天才能完成初步筛选。\n- **信息整合困难**：从不同文献中提取作者、观点和标签时，需反复切换窗口手动复制粘贴，极易出错。\n- **图表分析缺失**：遇到包含关键数据的复杂图表时，无法直接让 AI 解读图片内容，只能依靠肉眼估算或跳过。\n- **笔记整理繁琐**：将思考过程转化为正式笔记时，需要重新组织语言并手动关联原文位置，打断写作心流。\n\n### 使用 ai-research-assistant 后\n- **智能对话速读**：直接在 Zotero 侧边栏通过自然语言提问，ai-research-assistant 瞬间总结多篇文献的核心贡献与局限性。\n- **拖拽式引用交互**：利用拖拽功能将文献集合直接送入对话框，自动补全作者与标签，一键生成对比分析草稿。\n- **视觉深度解析**：启用 GPT-4 Vision 功能，直接框选 PDF 中的复杂图表，ai-research-assistant 立即输出数据趋势分析与文字解读。\n- **无缝笔记沉淀**：对话中的精彩洞察可一键保存为带原文锚点的注释和笔记，自动同步至 Zotero 条目，随时调用。\n\nai-research-assistant 将原本需要数天的文献梳理工作压缩至几小时，让研究者从繁琐的资料搬运工转变为真正的知识架构师。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Flifan0127_ai-research-assistant_c9c26f10.png","lifan0127","Fan Li","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Flifan0127_6642683d.jpg","Ph.D. Research Scientist","Apex 974 LLC","Wilmington, DE, United States",null,"FanLi_RnD","https:\u002F\u002Fapex974.com","https:\u002F\u002Fgithub.com\u002Flifan0127",[83,87,91,95,98,102],{"name":84,"color":85,"percentage":86},"JavaScript","#f1e05a",79.7,{"name":88,"color":89,"percentage":90},"TypeScript","#3178c6",19.7,{"name":92,"color":93,"percentage":94},"HTML","#e34c26",0.2,{"name":96,"color":97,"percentage":94},"Rust","#dea584",{"name":99,"color":100,"percentage":101},"CSS","#663399",0.1,{"name":103,"color":104,"percentage":101},"Fluent","#ffcc33",1684,116,"2026-04-09T03:16:17","AGPL-3.0","Windows, macOS, Linux","未说明",{"notes":112,"python":110,"dependencies":113},"该工具是 Zotero 插件，非独立运行的本地 AI 模型，因此无需本地 GPU、特定 Python 环境或显存。核心需求是安装对应版本的 Zotero（6 或 7），并配置有效的 OpenAI API Key。视觉分析功能需要 GPT-4 Vision 模型的预览访问权限。安装后需重启 Zotero 才能生效。",[114,115,116],"Zotero 6 或 Zotero 7","OpenAI API Key","GPT-4 模型系列访问权限",[118,35,13],"其他",[120,121,122,123,124],"ai-assistant","research-paper","zotero","gpt","large-language-models","2026-03-27T02:49:30.150509","2026-04-09T21:35:28.891533",[],[129,134,139,144,149,154,159,164,169,174,179,184,189,194,199,204,209,214,219,224],{"id":130,"version":131,"summary_zh":132,"released_at":133},171890,"v0.7.5","- 使用字体图标\n- 为 Zotero 7 更新字体大小","2024-10-20T17:54:26",{"id":135,"version":136,"summary_zh":137,"released_at":138},171891,"0.8.0","这是与 Zotero 6 兼容的最终版本。未来的版本将仅支持 Zotero 7。","2024-10-07T00:56:16",{"id":140,"version":141,"summary_zh":142,"released_at":143},171892,"v0.7.4","请升级 zotero-plugin-toolkit，以修复与其他插件的兼容性问题。\n\n请注意，安装完成后，您可能需要前往“工具”>“插件”，并启用该插件。\n\n![截图 2024-10-05 下午10:13:22](https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F0fa95330-ce63-437b-ac65-43bbedd3923c)","2024-10-06T01:57:49",{"id":145,"version":146,"summary_zh":147,"released_at":148},171893,"v0.7.2","- 将已弃用的 GPT-4-Vision 模型替换为用户专属模型","2024-09-13T02:27:53",{"id":150,"version":151,"summary_zh":152,"released_at":153},171894,"0.7.2","## 注意：此版本仅适用于 Zotero 6。对于 Zotero 7，请使用 https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002Fv0.7.4\n\n- 将已弃用的 GPT-4-Vision 模型替换为用户自定义模型","2024-09-13T02:14:53",{"id":155,"version":156,"summary_zh":157,"released_at":158},171895,"v0.7.1","- 添加对 GPT-4o 和 GPT-4o mini 的支持\n\n![图片](https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fa9a5c6f3-bc05-48f3-b7cd-9d519f77af82)\n","2024-08-29T02:28:43",{"id":160,"version":161,"summary_zh":162,"released_at":163},171896,"0.7.1","- 添加对 GPT-4o 和 GPT-4o mini 的支持\n\n![图片](https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Ff32271bc-4183-4d32-b6aa-6e3e201b4701)\n","2024-07-29T13:44:52",{"id":165,"version":166,"summary_zh":167,"released_at":168},171897,"0.7.0-z7","这是 Aria for Zotero 7 的实验性预发布版本（与 Zotero 6 不兼容）。它提供了与 0.7.0 版本相同的功能：https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Freleases\u002Ftag\u002F0.7.0","2024-04-15T12:30:44",{"id":170,"version":171,"summary_zh":172,"released_at":173},171898,"0.7.0","功能：\n- 交互式提示库\n- 常见问题解答\n\n![ARIA-0 7 0](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Fcc3cf022-39e9-4e82-82b5-36c2b32e4b39)\n","2024-01-24T14:18:18",{"id":175,"version":176,"summary_zh":177,"released_at":178},171899,"0.6.6","### 功能：\n- 可根据聊天消息创建 Zotero 笔记（包括独立笔记和子笔记）。\n- 可将聊天历史导出为 Zotero 独立笔记。\n- 可将视觉分析结果添加到 Zotero 图片标注中。\n\n![Aria-0 6 6-720p](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Fab20e024-322b-4b8b-b9a8-124ccd39a990)\n\n### 错误修复：\n- 修复自动补全高亮显示问题。\n- 修复超宽图片的图片标注显示问题。\n- 移除过时的发布说明。\n- 在更新聊天历史文件时处理 UTF-8 解码问题。","2023-12-24T00:13:13",{"id":180,"version":181,"summary_zh":182,"released_at":183},171900,"0.6.5","- Make Aria dialog resizable under macOS.\r\n- Update error handling for \"invalid_api_key\" and \"model_not_found\".\r\n\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002F087518a8-1fd8-4a9b-9dc0-b9c09e10fb08)\r\n\r\n\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Fe5562424-56eb-4f9c-a305-366fc9e27e34)\r\nS","2023-12-04T15:45:00",{"id":185,"version":186,"summary_zh":187,"released_at":188},171901,"0.6.4","Features:\r\n- Support custom OpenAI base URL (Must include the API version. e.g. \"\u002Fv1\")\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002F3a31d0c1-a3a1-455f-91ea-b79ad59d3a0d)\r\n\r\nBug Fixes:\r\n- Fix error related to saving chat history with large images\r\n- Various style fixes under dark mode.\r\n- Compatibility warning for certain Zotero 6 installations","2023-11-16T14:16:25",{"id":190,"version":191,"summary_zh":192,"released_at":193},171902,"0.6.3","- Open URL in your default browser","2023-11-13T04:25:15",{"id":195,"version":196,"summary_zh":197,"released_at":198},171903,"0.6.2","## Features\r\n\r\n- Persistent chat history after Zotero restart\r\n- Add Aria launch button in Zotero toolbar\r\n\r\n\r\n## Bug Fixes:\r\n- Search title only for item autocomplete\r\n- Fix zoom menu item background color in dark mode","2023-11-12T19:40:35",{"id":200,"version":201,"summary_zh":202,"released_at":203},171904,"0.6.1","- Allow customization of shortcut combo for Aria invocation\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002F3b669d66-782b-4a9f-ac97-476fd53952c3)\r\n\r\n- Allow zoom in\u002Fout for Aria window\r\n![zoom](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Faacc2839-ed27-4106-ae29-b8f03182c692)\r\n\r\n- Add relevant item content to vision chain\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Fff52e1ff-0b77-4089-b4f4-7323fdf5f2ee)\r\n","2023-11-11T19:58:11",{"id":205,"version":206,"summary_zh":207,"released_at":208},171905,"0.6.0","- Support visual analysis of graphics, diagrams, plots, tables and text via GPT-4 Vision.\r\n\r\n![0 6 0](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002Faa3d5d6f-817b-41e5-81a2-ba181ad9e0e5)\r\n\r\n\r\n","2023-11-09T16:05:29",{"id":210,"version":211,"summary_zh":212,"released_at":213},171906,"0.5.5","Add support for the new OpenAI GPT-4 Turbo model (gpt-4-1106-preview). Choose your preferred model under Preferences > Aria\r\n\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002F68f20ac1-e962-456c-9159-3cdb5487d2e1)\r\n\r\n\r\n","2023-11-07T22:19:56",{"id":215,"version":216,"summary_zh":217,"released_at":218},171907,"0.5.4","## Bug Fixes:\r\n\r\n- Add retry logic when fetching Zotero items (#61)","2023-11-05T15:15:24",{"id":220,"version":221,"summary_zh":222,"released_at":223},171908,"0.5.3","Should your interaction with Aria encounter any error, Aria will display the underlying Zotero error message to facilitate troubleshooting. \r\n\r\nBe aware, though, that this detailed error information might include personal data, specifically the path to the Zotero user profile directory where Aria is installed.\r\n\r\nThe intent of this patch release is to protect user privacy by masking such information when displayed in Aria or submitted through the feedback channel.","2023-11-04T22:50:08",{"id":225,"version":226,"summary_zh":227,"released_at":228},171909,"0.5.1","- Allow users to submit feedback and share chat session for Aria improvement.\r\n- Make copy and edit message buttons more prominent.\r\n\r\n![image](https:\u002F\u002Fgithub.com\u002Flifan0127\u002Fai-research-assistant\u002Fassets\u002F7218193\u002F479bd523-e0f7-45ad-9c78-da40b7abd676)\r\n\r\n\r\n","2023-11-03T16:52:06"]