[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"home-initial":3},{"overview":4,"hot":29,"featured":294,"langs":494,"topics":514,"newest":534},{"totalProjects":5,"curatedProjects":5,"totalAwesomeProjects":6,"totalStars":7,"totalForks":8,"todayNewProjects":9,"todayNewStars":10,"todayTrendingProjects":11,"totalLanguages":12,"totalTopics":13,"topLanguages":14,"fastestGrowingProject":20},17288,0,159047751,24181579,43,104421,187,95,2234,{"TypeScript":15,"Java":16,"JavaScript":17,"Go":18,"Python":19},3056,681,1472,1103,3902,{"id":21,"name":22,"fullName":23,"language":24,"stars":25,"stars1d":26,"htmlUrl":27,"description":28},80780,"defending-code-reference-harness","anthropics\u002Fdefending-code-reference-harness","Python",2981,2165,null,"Skills for threat modeling, scanning, triage, patching, plus an autonomous scanning harness you can \u002Fcustomize",[30,65,92,124,145,164,192,219,252,280],{"id":31,"name":32,"fullName":33,"owner":34,"description":35,"aiSummary":36,"language":24,"stars":37,"forks":38,"stars7d":39,"stars30d":40,"starsTrendScore":41,"htmlUrl":27,"topics":42,"license":58,"trendingRank":59,"starsGained":60,"trendingDate":61,"trendingSlot":64},290,"hermes-agent","NousResearch\u002Fhermes-agent","NousResearch","The agent that grows with you","Hermes Agent 是一个由Nous Research开发的自我改进型AI代理。它具有内置的学习循环，能够从经验中创建技能，并在使用过程中不断优化这些技能，同时还能搜索过往对话并构建用户模型。该项目支持多种语言模型，如OpenAI、Hugging Face等，并且可以在低成本VPS或云基础设施上运行。其核心功能包括全终端界面、跨平台消息传递、自动技能提升以及定时任务自动化。适合需要持续学习和适应特定用户需求的各种场景，例如个人助理、客户服务机器人或是任何需要智能交互的应用。",182712,31325,9964,45311,9485,[43,44,45,46,47,48,49,50,51,52,32,53,54,55,56,57],"ai","ai-agent","ai-agents","anthropic","chatgpt","claude","claude-code","clawdbot","codex","hermes","llm","moltbot","nous-research","openai","openclaw","MIT License",1,1845,[62,63,63],2026,6,"daily",{"id":66,"name":67,"fullName":68,"owner":69,"description":70,"aiSummary":71,"language":24,"stars":72,"forks":73,"stars7d":74,"stars30d":75,"starsTrendScore":6,"htmlUrl":27,"topics":76,"license":88,"trendingRank":89,"starsGained":90,"trendingDate":91,"trendingSlot":64},74213,"headroom","chopratejas\u002Fheadroom","chopratejas","Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.","Headroom 是一个专为大型语言模型（LLM）应用设计的上下文优化层。它通过压缩AI代理读取的所有内容，如工具输出、日志、检索增强生成（RAG）片段、文件和对话历史记录，在保持相同答案质量的同时显著减少了所需的令牌数量。项目支持多种算法，具备本地优先且可逆的特点，并提供库、代理及代理包装器等多种使用方式，适用于需要高效处理大量文本数据以降低成本或提高性能的各种AI应用场景。",14352,912,690,1027,[77,43,46,78,79,80,81,82,53,83,56,84,85,86,87],"agent","compression","context-engineering","context-window","fastapi","langchain","mcp","proxy","python","rag","token-optimization","Apache License 2.0",2,2473,[62,63,63],{"id":93,"name":94,"fullName":95,"owner":94,"description":96,"aiSummary":97,"language":98,"stars":99,"forks":100,"stars7d":101,"stars30d":102,"starsTrendScore":103,"htmlUrl":27,"topics":104,"license":58,"trendingRank":121,"starsGained":122,"trendingDate":123,"trendingSlot":64},3585,"CopilotKit","CopilotKit\u002FCopilotKit","The Frontend Stack for Agents & Generative UI. React + Angular.  Makers of the AG-UI Protocol","CopilotKit 是一个用于构建全栈代理应用、生成式UI和聊天应用的顶级SDK。它支持React和Angular，核心功能包括基于React的聊天界面、后端工具渲染、动态生成UI组件、共享状态以及人机协作工作流。该项目特别适合需要集成AI助手或生成式用户界面的应用场景，如智能客服系统、个性化推荐引擎等。通过采用AG-UI协议，CopilotKit能够确保与多个主流技术平台的良好兼容性。","TypeScript",32543,4185,659,1763,1385,[77,105,106,107,43,44,108,109,110,111,112,113,114,53,115,116,117,118,119,120],"agent-native","agentic-ai","agents","ai-assistant","assistant","assistant-chat-bots","copilot","copilot-chat","generative-ui","js","nextjs","open-source","react","reactjs","ts","typescript",3,366,[62,63,63],{"id":125,"name":126,"fullName":127,"owner":128,"description":129,"aiSummary":130,"language":98,"stars":131,"forks":132,"stars7d":133,"stars30d":134,"starsTrendScore":135,"htmlUrl":27,"topics":136,"license":58,"trendingRank":142,"starsGained":143,"trendingDate":144,"trendingSlot":64},3683,"open-notebook","lfnovo\u002Fopen-notebook","lfnovo","An Open Source implementation of Notebook LM with more flexibility and features","Open Notebook 是一个开源的笔记应用，提供了比传统笔记应用更灵活和丰富的功能。它支持多种AI模型选择（包括OpenAI、Anthropic等18种以上），能够处理多模态内容如PDF、视频、音频及网页，并具备生成专业播客、智能搜索以及基于上下文的AI对话等功能。此外，该应用注重用户隐私保护，所有数据均本地存储。适用于需要高度定制化学习环境和个人知识管理的场景，特别适合研究人员、学生以及终身学习者使用。",25832,2973,1882,2688,3858,[109,137,138,139,140,141],"learning","note-taking","notebook","notes-app","self-learning",4,1152,[62,63,63],{"id":146,"name":147,"fullName":148,"owner":149,"description":150,"aiSummary":151,"language":152,"stars":153,"forks":154,"stars7d":155,"stars30d":156,"starsTrendScore":157,"htmlUrl":27,"topics":158,"license":58,"trendingRank":161,"starsGained":162,"trendingDate":163,"trendingSlot":64},77592,"ECC","affaan-m\u002FECC","affaan-m","The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.","ECC 是一个针对代理性能优化的系统，旨在提升Claude Code、Codex、Cursor等AI代理工具的工作效率。该项目通过集成技能、直觉、记忆优化、持续学习和安全扫描等功能，提供了一套完整的解决方案，支持多种语言生态系统。其核心在于为开发者提供生产就绪的代理、技能配置及规则设定，特别适合需要高度定制化与高效协作的软件开发场景。采用MIT许可协议，社区活跃度高，已获得广泛认可。","JavaScript",208220,31952,9791,20513,7241,[45,46,48,49,159,53,83,160],"developer-tools","productivity",5,1361,[62,63,63],{"id":165,"name":166,"fullName":167,"owner":168,"description":169,"aiSummary":170,"language":24,"stars":171,"forks":172,"stars7d":173,"stars30d":174,"starsTrendScore":175,"htmlUrl":27,"topics":176,"license":58,"trendingRank":63,"starsGained":190,"trendingDate":191,"trendingSlot":64},2538,"Agent-Reach","Panniantong\u002FAgent-Reach","Panniantong","Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.","Agent Reach 是一个为 AI 代理提供互联网访问能力的工具，支持读取和搜索 Twitter、Reddit、YouTube、GitHub、Bilibili 和小红书等多个平台的内容。其核心功能包括无需 API 费用即可通过命令行界面（CLI）实现跨平台的数据抓取与解析，利用开源工具如 yt-dlp 和 rdt-cli 等进行内容提取，并且能够自动处理网页结构化数据，使得 AI 可以直接理解文本信息而非原始 HTML 代码。此项目特别适用于需要增强 AI 代理的信息获取能力但又希望避免复杂配置或高昂成本的场景，比如自动化社交媒体分析、在线教育资源整合等。同时，Agent Reach 还保证了用户隐私安全，所有操作都在本地完成，不涉及敏感信息上传。",21460,1856,868,2557,1083,[177,44,178,179,180,49,181,182,183,184,83,85,185,186,187,188,189],"agent-infrastructure","ai-search","automation","bilibili","cli","cursor","free-api","llm-tools","reddit-scraper","twitter-scraper","web-scraper","xiaohongshu","youtube-transcript",148,[62,63,63],{"id":193,"name":194,"fullName":195,"owner":196,"description":197,"aiSummary":198,"language":24,"stars":199,"forks":200,"stars7d":201,"stars30d":202,"starsTrendScore":203,"htmlUrl":27,"topics":204,"license":215,"trendingRank":216,"starsGained":217,"trendingDate":218,"trendingSlot":64},545,"MiroFish","666ghj\u002FMiroFish","666ghj","A Simple and Universal Swarm Intelligence Engine, Predicting Anything. 简洁通用的群体智能引擎，预测万物","MiroFish 是一个基于多智能体技术的下一代AI预测引擎，能够通过提取现实世界中的种子信息（如突发新闻、政策草案或金融信号）自动构建高保真度的平行数字世界。其核心功能包括利用具有独立个性、长期记忆和行为逻辑的数千个智能代理进行自由互动与社会演化，并支持从“上帝视角”动态注入变量以精确推演未来趋势。该项目采用Python语言开发，具备强大的代理记忆、知识图谱以及多智能体模拟等技术特点，适用于需要对未来场景进行预测分析的各种场合，如金融预测、公共舆论分析和社会预测等领域。",64604,10078,1433,5216,1565,[205,206,207,208,209,210,211,212,213,214],"agent-memory","financial-forecasting","future-prediction","knowledge-graph","llms","multi-agent-simulation","public-opinion-analysis","python3","social-prediction","swarm-intelligence","GNU Affero General Public License v3.0",8,320,[62,63,63],{"id":220,"name":221,"fullName":222,"owner":223,"description":224,"aiSummary":225,"language":24,"stars":226,"forks":227,"stars7d":228,"stars30d":229,"starsTrendScore":230,"htmlUrl":27,"topics":231,"license":58,"trendingRank":249,"starsGained":250,"trendingDate":251,"trendingSlot":64},2250,"last30days-skill","mvanhorn\u002Flast30days-skill","mvanhorn","AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary","\u002Flast30days-skill 是一个AI代理技能，能够跨Reddit、X、YouTube、HN、Polymarket和网络研究任何主题，并生成基于实际互动的总结。其核心功能包括利用多个平台的数据（如点赞、评论、视频字幕等）进行深度搜索与分析，并通过AI代理综合这些信息生成简洁的报告。该项目采用Python语言编写，具有无需配置即可快速部署的特点，支持多种社交媒体及新闻网站的内容检索。特别适合需要紧跟最新趋势、热点事件或个人动态的研究者、分析师以及对特定领域保持关注的专业人士使用。",28125,2388,1322,3192,2492,[232,233,234,48,49,235,236,237,238,57,239,240,241,242,243,244,245,246,247,248],"ai-prompts","ai-skill","bluesky","clawhub","deep-research","hackernews","instagram","polymarket","recency","reddit","research","social-media","tiktok","trends","twitter","web-search","youtube",9,731,[62,63,63],{"id":253,"name":254,"fullName":255,"owner":256,"description":257,"aiSummary":258,"language":24,"stars":259,"forks":260,"stars7d":261,"stars30d":262,"starsTrendScore":263,"htmlUrl":27,"topics":264,"license":88,"trendingRank":277,"starsGained":278,"trendingDate":279,"trendingSlot":64},406,"PaddleOCR","PaddlePaddle\u002FPaddleOCR","PaddlePaddle","Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images\u002FPDFs and LLMs. Supports 100+ languages.","PaddlePaddle\u002FPaddleOCR 是一个强大的轻量级 OCR 工具包，能够将 PDF 或图像文档转换为结构化数据，适用于 AI 应用。该项目支持超过 100 种语言，并且集成了多种功能，如文档解析、翻译和关键信息提取等。其核心技术基于 PaddlePaddle 深度学习框架，具备高性能与灵活性，支持 CPU、GPU 等多种硬件加速。PaddleOCR 适合需要从非结构化文本中提取信息的场景，例如文档自动化处理、多语言内容分析以及与其他大型语言模型集成的应用。",80452,10628,1473,3247,2522,[265,266,267,268,269,270,271,272,273,274,275,276,86],"ai4science","chineseocr","document-parsing","document-translation","kie","ocr","paddleocr-vl","pdf-extractor-rag","pdf-parser","pdf2markdown","pp-ocr","pp-structure",10,747,[62,63,63],{"id":281,"name":282,"fullName":283,"owner":56,"description":284,"aiSummary":285,"language":152,"stars":286,"forks":287,"stars7d":288,"stars30d":289,"starsTrendScore":6,"htmlUrl":290,"topics":27,"license":27,"trendingRank":291,"starsGained":292,"trendingDate":293,"trendingSlot":64},11201,"plugins","openai\u002Fplugins","OpenAI Plugins","openai\u002Fplugins 是一个收集了Codex插件示例的项目。该项目通过提供一系列插件来增强开发者的生产力，每个插件都包含在`plugins\u002F\u003Cname>\u002F`目录下，并且必须有一个`.codex-plugin\u002Fplugin.json`配置文件以及可选的支持文件如技能、命令等。技术特点在于其支持多种应用场景下的自动化任务处理，比如使用Figma进行设计系统规则管理、Notion中的知识管理和会议记录、构建iOS\u002FmacOS\u002FWeb应用时的工作流优化等。适用于需要提高软件开发效率或希望集成特定工具到自己工作流程中的开发者。",1493,237,32,295,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fplugins",11,49,[62,63,63],[295,309,319,350,363,375,392,401,423,448,468,483],{"id":296,"name":297,"fullName":298,"owner":299,"description":300,"aiSummary":301,"language":24,"stars":302,"forks":303,"stars7d":304,"stars30d":305,"starsTrendScore":306,"rankGlobal":27,"rankLanguage":27,"license":88,"topics":307,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},76164,"minWM","shengshu-ai\u002FminWM","shengshu-ai","A Minimal and Elegant Framework & Tutorial for Real-Time Interactive World Models","minWM 是一个用于实时交互世界模型的全栈开源框架与教程。它提供了一个从数据准备到训练再到推理的完整流程，支持多骨干网络和多条件注入，并且通过Claude技能让使用者能够借助大语言模型助手修改框架。此外，项目还特别为新手提供了入门知识，帮助他们更好地理解和使用世界模型。适用于希望快速上手并深入研究世界模型的研究者和开发者，尤其是在需要构建基于动作条件的视频世界模型时。",541,7,241,424,60,[],false,{"id":310,"name":311,"fullName":312,"owner":313,"description":27,"aiSummary":314,"language":315,"stars":316,"forks":317,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":318,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},7606,"sanogueralorenzo.github.io","sanogueralorenzo\u002Fsanogueralorenzo.github.io","sanogueralorenzo","该项目是一个以AI工具为核心的编排实验平台。它主要由Codex Core、Codex Remote、Codex Menubar和Codex Web等组件构成，使用Kotlin语言开发，通过明确的合约、确定性检查及可恢复的工作流来实现高效的任务调度与执行。特别适合需要集成多种AI服务并进行复杂任务协调的应用场景，如自动化测试环境搭建、持续集成\u002F持续部署流程优化等。项目采用MIT许可证开源，社区活跃度高，便于开发者学习和贡献代码。","Kotlin",1787,326,[],{"id":320,"name":321,"fullName":322,"owner":323,"description":324,"aiSummary":325,"language":326,"stars":190,"forks":59,"stars7d":327,"stars30d":328,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":329,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},78859,"tomodachi-pc","RayceAnderson\u002Ftomodachi-pc","RayceAnderson","tomodachi life pc living  the dream windows  free patch notes update ryujinx emulator rom nsp xci mii creator life simulator 60fps mod graphics config save file setup guide fix crash","Tomodachi PC 项目通过 Ryujinx 模拟器将 Nintendo Switch 版本的 Tomodachi Life 游戏移植到 Windows 平台，提供全功能支持。其核心功能包括60FPS流畅体验、高分辨率显示、Mii角色创建与编辑、存档管理以及图形和性能优化，并且支持模组。该项目适合拥有合法游戏副本并希望在PC上享受更高画质和更流畅游戏体验的玩家使用。此外，它还提供了简易安装指南和定期更新，确保用户能够方便地设置和维护游戏环境。","C#",121,343,[330,331,332,333,334,335,336,337,338,339,340,341,342,343,344,345,346,347,348,349],"\"[life-sim","life-simulation","mii","mii-sharing","nes-emulator","nintendo","nintendo-64","nintendo-ds-emulator","nintendo-port","nintendo-switch","nintendo-switch-emulator","nintendo-tomodachi-life","tomodachi","tomodachi-collection","tomodachi-life","tomodachi-life-desktop","tomodachi-life-living-the-dream","tomodachi-life-pc","tomodachi-life-windows","tomodachi-living-the-dream]\"",{"id":351,"name":352,"fullName":353,"owner":354,"description":355,"aiSummary":356,"language":152,"stars":357,"forks":358,"stars7d":359,"stars30d":360,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":361,"topics":362,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},74866,"dbskill","dontbesilent2025\u002Fdbskill","dontbesilent2025","dontbesilent 的商业诊断 Skills","dontbesilent2025\u002Fdbskill 是一个商业诊断工具箱，通过从大量推文中提炼的方法论形成18个Agent技能，帮助用户在多个支持skill或system prompt的平台上进行商业分析与个人成长规划。该项目采用Shell脚本编写，核心功能包括商业模式诊断、对标分析、内容创作优化等，并特别强调了交互式学习和目标清晰化工具，能够根据用户反馈调整学习材料，将模糊目标转化为具体可执行的任务。此外，还提供了诊断状态管理功能，支持诊断过程的保存、恢复及报告生成，适合需要系统性解决商业问题、提升个人能力或项目管理水平的场景使用。",6199,813,503,1166,"Other",[],{"id":364,"name":365,"fullName":366,"owner":367,"description":368,"aiSummary":369,"language":24,"stars":370,"forks":371,"stars7d":372,"stars30d":373,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":374,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},74023,"OpenSpace","HKUDS\u002FOpenSpace","HKUDS","\"OpenSpace: Make Your Agents: Smarter, Low-Cost, Self-Evolving\" -- Community: https:\u002F\u002Fopen-space.cloud\u002F","OpenSpace 是一个旨在让AI代理变得更智能、成本更低且能够自我进化的开源项目。它通过减少46%的令牌使用量和实现技能自我进化，帮助用户节省成本并提升效率。项目支持多种主流AI代理如Claude Code、Codex等，并提供一站式的命令行工具来管理和优化这些代理的行为。此外，OpenSpace还具备多渠道通信网关功能，允许与外部平台如WhatsApp和Feishu进行交互。适用于需要构建高效、低成本且能够持续自我优化的人工智能解决方案的场景，特别适合开发者或团队希望增强其AI应用性能的情况。",6426,795,18,267,[],{"id":376,"name":377,"fullName":378,"owner":379,"description":380,"aiSummary":381,"language":382,"stars":383,"forks":384,"stars7d":385,"stars30d":386,"starsTrendScore":387,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":388,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},4796,"beszel","henrygd\u002Fbeszel","henrygd","Lightweight server monitoring with historical data, docker stats, and alerts.","Beszel 是一个轻量级的服务器监控平台，提供Docker统计、历史数据记录和警报功能。它采用Go语言编写，具有用户友好的Web界面和简单的配置流程，开箱即用。该平台支持自动备份、多用户管理、OAuth认证以及API访问等功能。其核心组件包括用于展示和管理系统信息的Hub和安装在被监控系统上负责收集并发送性能指标的Agent。由于其低资源消耗特性和易用性，Beszel非常适合家庭实验室环境和个人或小型团队维护自托管服务时使用。","Go",22435,824,222,790,141,[389,390,391],"[\\\"homelab\\","\\\"monitoring\\\"","\\\"self-hosted\\\"]\"",{"id":393,"name":394,"fullName":395,"owner":394,"description":396,"aiSummary":397,"language":27,"stars":398,"forks":399,"stars7d":89,"stars30d":121,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":27,"topics":400,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},71494,"coder2gwy","coder2gwy\u002Fcoder2gwy","互联网首份程序员考公指南，由3位已经进入体制内的前大厂程序员联合献上。","该项目是一份由三位已成功进入体制内的前大厂程序员联合编写的考公指南，旨在帮助希望转行至公务员、事业单位或教师岗位的程序员。核心功能包括分享备考经验、提供实用建议及解答常见问题，技术特点在于其基于真实经历的指导性和实用性。适合于那些经过深思熟虑后认为当前程序员职业路径无法满足个人需求，并考虑转向更加稳定工作环境的IT从业者参考使用。",27587,3706,[],{"id":402,"name":403,"fullName":404,"owner":405,"description":406,"aiSummary":407,"language":27,"stars":408,"forks":409,"stars7d":410,"stars30d":411,"starsTrendScore":412,"rankGlobal":27,"rankLanguage":27,"license":215,"topics":413,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},1730,"Clone-Wars","GorvGoyl\u002FClone-Wars","GorvGoyl","100+ open-source clones of popular sites like Airbnb, Amazon, Instagram, Netflix, Tiktok, Spotify, Whatsapp, Youtube etc. See source code, demo links, tech stack, github stars.","Clone Wars 是一个集合了超过100个流行网站开源克隆项目的资源库，如Airbnb、Amazon、Instagram等。该项目不仅提供了源代码，还附带了演示链接、技术栈信息以及GitHub星标数量，非常适合开发者学习和参考。其核心功能包括提供全栈克隆项目及其教程链接，分为UI相似但非完全功能的克隆版与完全功能的开源替代品两大类。无论是希望深入理解现代Web开发技术栈，还是寻找灵感来构建自己的应用，Clone Wars都是一个极佳的选择。",34262,3073,22,104,23,[414,415,416,417,418,419,420,421,422],"[\\\"airbnb\\","\\\"clone\\\"","\\\"instagram\\\"","\\\"netflix\\\"","\\\"reddit\\\"","\\\"spotify\\\"","\\\"tiktok\\\"","\\\"twitter\\\"","\\\"whatsapp\\\"]\"",{"id":424,"name":425,"fullName":426,"owner":427,"description":428,"aiSummary":429,"language":430,"stars":431,"forks":432,"stars7d":433,"stars30d":434,"starsTrendScore":435,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":436,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},820,"ruff","astral-sh\u002Fruff","astral-sh","An extremely fast Python linter and code formatter, written in Rust.","Ruff 是一个用 Rust 编写的极快速的 Python 代码检查工具和格式化器。它比现有的代码检查工具（如 Flake8）和格式化工具（如 Black）快 10 到 100 倍，并且支持通过 `pip` 安装、`pyproject.toml` 配置以及 Python 3.14 兼容性。Ruff 内置了超过 900 条规则，能够自动修复错误（例如删除未使用的导入），并且支持缓存以避免重复分析未更改的文件。此外，Ruff 还提供了与 VS Code 等编辑器的集成，非常适合用于大型项目或需要高效代码检查和格式化的开发环境。","Rust",47829,2128,99,427,91,[437,438,439,440,441,442,443,444,445,446,447],"[\\\"linter\\","\\\"pep8\\\"","\\\"python\\\"","\\\"python3\\\"","\\\"ruff\\\"","\\\"rust\\\"","\\\"rustpython\\\"","\\\"static-analysis\\\"","\\\"static-code-analysis\\\"","\\\"style-guide\\\"","\\\"styleguide\\\"]\"",{"id":449,"name":450,"fullName":451,"owner":452,"description":453,"aiSummary":454,"language":24,"stars":455,"forks":456,"stars7d":372,"stars30d":457,"starsTrendScore":458,"rankGlobal":27,"rankLanguage":27,"license":88,"topics":459,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},10214,"gorilla","ShishirPatil\u002Fgorilla","ShishirPatil","Gorilla: Training and Evaluating LLMs for Function Calls (Tool Calls)","Gorilla 是一个用于训练和评估大型语言模型以执行函数调用（工具调用）的项目。它通过连接大量的API，使语言模型能够更好地理解和使用外部工具。该项目利用Python编写，支持多种主流的LLM API如OpenAI、Claude等，并且提供了一个详细的基准测试平台来比较不同模型在处理复杂任务时的表现，包括多步推理、错误恢复以及格式敏感性等方面的能力。适用于需要增强现有语言模型功能的应用场景，比如构建更智能的聊天机器人或自动化助手。",12893,1376,36,13,[460,461,462,463,464,465,466,467],"[\\\"api\\","\\\"api-documentation\\\"","\\\"chatgpt\\\"","\\\"claude-api\\\"","\\\"gpt-4-api\\\"","\\\"llm\\\"","\\\"openai-api\\\"","\\\"openai-functions\\\"]\"",{"id":469,"name":470,"fullName":471,"owner":472,"description":473,"aiSummary":474,"language":98,"stars":475,"forks":476,"stars7d":59,"stars30d":161,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":477,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},73844,"AingDesk","aingdesk\u002FAingDesk","aingdesk","AingDesk是一款简单好用的AI助手，支持知识库、模型API、分享、联网搜索、智能体，它还在飞快成长中。 AingDesk is a simple and easy-to-use AI assistant that supports knowledge bases, model APIs, sharing, internet search, and intelligent agents. It is still growing rapidly.","AingDesk是一款简单易用的AI助手，支持知识库、模型API、分享、联网搜索和智能体功能。其核心功能包括一键部署本地AI模型及主流模型API、创建本地知识库、构建可共享的智能体以及实现网页搜索等。此外，AingDesk还支持服务器端部署，并提供MCP客户端版本。这款工具非常适合需要快速搭建AI应用的开发者或企业使用，在日常办公、教育学习等多个场景下都能发挥重要作用。",2511,280,[478,479,465,480,481,482],"[\\\"deepseek\\","\\\"electron\\\"","\\\"localai\\\"","\\\"nodejs\\\"","\\\"ollama\\\"]\"",{"id":484,"name":485,"fullName":486,"owner":487,"description":27,"aiSummary":488,"language":382,"stars":489,"forks":490,"stars7d":491,"stars30d":492,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":58,"topics":493,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},74655,"whitelist-bypass","kulikov0\u002Fwhitelist-bypass","kulikov0","该项目通过视频通话平台（如VK Call、Yandex Telemost、WB Stream）隧道传输互联网流量，以绕过政府的白名单审查。其核心功能包括两种隧道模式：基于SCTP的数据通道（DC）和基于VP8视频编码的视频模式。项目采用Go语言编写，并利用Pion库实现端到端通信，支持ChaCha20加密和可配置的VP8速率控制。推荐使用无头模式部署，直接与平台的SFU交互，无需浏览器参与。适合在需要绕过网络审查限制的情况下使用，尤其是当常规方法受限时。",1373,75,34,304,[],[495,496,497,498,500,501,503,506,509,512],{"language":24,"projectCount":19,"avgStars":27,"totalStars":27,"rank":59},{"language":152,"projectCount":17,"avgStars":27,"totalStars":27,"rank":89},{"language":98,"projectCount":15,"avgStars":27,"totalStars":27,"rank":121},{"language":499,"projectCount":16,"avgStars":27,"totalStars":27,"rank":142},"Java",{"language":382,"projectCount":18,"avgStars":27,"totalStars":27,"rank":161},{"language":430,"projectCount":502,"avgStars":27,"totalStars":27,"rank":63},972,{"language":504,"projectCount":505,"avgStars":27,"totalStars":27,"rank":303},"C++",456,{"language":507,"projectCount":508,"avgStars":27,"totalStars":27,"rank":216},"C",682,{"language":510,"projectCount":511,"avgStars":27,"totalStars":27,"rank":249},"Swift",640,{"language":315,"projectCount":513,"avgStars":27,"totalStars":27,"rank":277},554,[515,517,519,521,524,526,528,529,531,532],{"topic":516,"projectCount":303,"totalStars":27,"rank":59},"machine-learning",{"topic":518,"projectCount":161,"totalStars":27,"rank":89},"deep-learning",{"topic":520,"projectCount":142,"totalStars":27,"rank":121},"artificial-intelligence",{"topic":522,"projectCount":523,"totalStars":27,"rank":142},"kubernetes",12,{"topic":525,"projectCount":523,"totalStars":27,"rank":161},"docker",{"topic":527,"projectCount":277,"totalStars":27,"rank":63},"vue",{"topic":117,"projectCount":303,"totalStars":27,"rank":303},{"topic":530,"projectCount":303,"totalStars":27,"rank":216},"nodejs",{"topic":85,"projectCount":410,"totalStars":27,"rank":249},{"topic":533,"projectCount":303,"totalStars":27,"rank":277},"api",[535,550,560,574,587],{"id":536,"name":537,"fullName":538,"owner":56,"description":539,"aiSummary":27,"language":510,"stars":540,"forks":541,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":361,"topics":542,"htmlUrl":27,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":549},83359,"tart","openai\u002Ftart","macOS and Linux VMs on Apple Silicon to use in CI and other automations",5765,294,[543,179,544,545,546,537,547,548],"apple-silicon","ci","fair-source","macos","virtualization","virtualization-framework","2026-06-05 17:13",{"id":551,"name":552,"fullName":553,"owner":554,"description":555,"aiSummary":27,"language":504,"stars":556,"forks":557,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":27,"topics":27,"htmlUrl":558,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":559},83310,"workerd","cloudflare\u002Fworkerd","cloudflare","The JavaScript \u002F Wasm runtime that powers Cloudflare Workers",8280,649,"https:\u002F\u002Fgithub.com\u002Fcloudflare\u002Fworkerd","2026-06-05 11:47",{"id":561,"name":562,"fullName":563,"owner":564,"description":565,"aiSummary":27,"language":504,"stars":566,"forks":567,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":27,"topics":568,"htmlUrl":573,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},83313,"kalibr","ethz-asl\u002Fkalibr","ethz-asl","The Kalibr visual-inertial calibration toolbox",5487,1574,[569,570,571,572],"camera","calibration","calibration-toolbox","imu","https:\u002F\u002Fgithub.com\u002Fethz-asl\u002Fkalibr",{"id":575,"name":576,"fullName":577,"owner":578,"description":579,"aiSummary":27,"language":504,"stars":580,"forks":581,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":27,"topics":582,"htmlUrl":586,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":27},83309,"LiteRT-LM","google-ai-edge\u002FLiteRT-LM","google-ai-edge","LiteRT-LM is Google's production-ready, high-performance, open-source inference framework for deploying Large Language Models on edge devices.",5420,555,[583,584,585],"edge-ai","on-device-ai","on-device-llm","https:\u002F\u002Fgithub.com\u002Fgoogle-ai-edge\u002FLiteRT-LM",{"id":588,"name":589,"fullName":590,"owner":591,"description":592,"aiSummary":27,"language":504,"stars":593,"forks":594,"stars7d":6,"stars30d":6,"starsTrendScore":6,"rankGlobal":27,"rankLanguage":27,"license":27,"topics":27,"htmlUrl":595,"archived":308,"starSnapshotCount":6,"trendingCount":6,"createdAt":27,"pushedAt":596},83314,"NeuralAmpModelerPlugin","sdatkinson\u002FNeuralAmpModelerPlugin","sdatkinson","Plugin for Neural Amp Modeler",2706,220,"https:\u002F\u002Fgithub.com\u002Fsdatkinson\u002FNeuralAmpModelerPlugin","2026-06-02 19:21"]