GitHub Trending 2026-11-13 每日热门项目榜单
2026年11月13日 GitHub Trending 共 50 个热门开源项目,按当日排名排列。
AI 日报
AI Coding Agent 生态裂变:大厂规则注入与工程技能包成为新基建
2026 年 11 月 13 日 · GitHub Trending 每日观察
今日榜单清晰展现了 AI 辅助编程领域从“单一对话生成”向“Agent 工程协同与执行基建”的全面演进。Google Chrome 团队与 OpenAI 亲自下场,通过规范指导与自动化安全检测为编程 Agent 设立行业准则。与此同时,Agent Skills 插件生态与低成本分发体系呈现爆发态势,社区关注点已迅速延伸至上下文优化、审美控制与自动化代码审查等纵深领域。
今日数据速览
| 指标 | 数值 | 解读 | |---|---|---| | 单日星增榜首 | GoogleChrome/modern-web-guidance-src +1,875 | 官方团队为 Agent 制定 Web 标准 | | 个人财务黑马 | we-promise/sure +944 | 开源记账应用凭借协作理念逆势走红 | | 开发者生态热度最高 | Alishahryar1/free-claude-code +895 | 多 Agent 免费调度方案需求极为旺盛 | | 上榜项目最高总星数 | affaan-m/ECC 302,439 | Agent 性能优化底座积累庞大用户群 | | 经典推理基石持续增长 | ggml-org/llama.cpp +613 | 本地高效推理仍然是不可替代的刚需 |
单日增星前列被官方规范与 Agent 增强工具占据,其中 Google Chrome 团队的现代 Web 指南单日吸星 1,875 颗拔得头筹,体现了开发者对 AI 生成代码符合现代 Web 最佳实践的强烈诉求。
热门趋势项目详解
1. GoogleChrome/modern-web-guidance-src — Agent 专用的现代 Web 最佳实践库
TypeScript · 4,759 星 · 单日 +1,875
该项目由 Google Chrome 团队维护,旨在直接向 AI 编码智能体注入最新的现代 Web 标准与开发规范,解决大模型常因训练数据滞后而生成过时 HTML/CSS/JS 代码的问题。随着各大编码 Agent 深入前端工作流,缺乏更新标准的 Agent 往往会引入陈旧的反模式代码。该仓库通过结构化、可被 Agent 快速解析的最佳实践指南,使智能体能够对齐当下最新的浏览器特性与性能优化标准。这一举措极大提升了前端代码的生成质量,但开发者仍需警惕 Agent 过度依赖新特性可能引发的旧版浏览器兼容性挑战。
2. we-promise/sure — 社区驱动的开源个人财务应用
Ruby · 14,697 星 · 单日 +944
在技术基建唱主角的榜单中,这款纯粹面向终端用户的个人记账软件脱颖而出。它主打人人可用、人人参与构建的开源协作理念,为注重隐私与数据自主权的用户提供了抗衡商业闭源记账产品的替代品。该项目今日爆火,一方面源于社区对跨平台统一、数据自持财务工具的长期渴望,另一方面也反映出 Ruby 社区在快速交付成熟 Web 应用上的深厚积淀。不过作为涉及敏感资产数据的软件,社区能否持续保障多方贡献代码的安全性与合规性,将是其未来面临的关键考验。
3. Alishahryar1/free-claude-code — 多端免费 Agent 路由与运行套件
Python · 69,867 星 · 单日 +895
该项目提供了一整套跨终端、跨 IDE 及浏览器的 Agent 运行解决方案,支持 Claude Code、Codex、Antigravity 等多种主流 Agent Harness。据项目自述,其集成了超过 13 亿免费 Token 的调度支持,并提供了原生浏览器会话及语音交互能力。它直击当前开发者多头采购 AI API 成本高昂、不同厂商客户端割裂的痛点,迅速积累了近 7 万星。但必须指出,聚合第三方接口与免费额度的方案通常伴随着服务稳定性不可控以及 API 服务商服务条款(ToS)调整的连带风险。
4. CopilotKit/OpenBot — 拥有独立沙箱环境的开源 AI 同事
TypeScript · 12,096 星 · 单日 +799
该项目打破了传统代码补全插件的局限,为每个 AI 协作者配置了专属虚拟计算机,内含浏览器、文件系统与工具集。其核心机制在于实现行为的“事前决策与事后审计”,能够兼容任意 AG-UI 智能体。今日受到追捧的原因在于,开发者日益厌倦只能停留在文本框层面的 AI 工具,需要真正能够执行完整交付任务的“自主同事”。这种赋予 Agent 独立操作环境的设计大大拓宽了自动化的边界,但其运行隔离安全性与误操作成本控制,仍需要更严格的企业级环境检验。
5. Leonxlnx/taste-skill — 赋予编码 Agent 审美与设计品味
JavaScript · 115,016 星 · 单日 +783
长久以来,AI 独立生成的网页常被诟病具有千篇一律的“AI 塑料感”。该项目通过引入艺术指导规则、参考图像解析以及自动化浏览器视觉检查机制,训练 Agent 生成具备现代审美质感的高级界面,而不是机械拼接通用组件。这一项目在前端社区引发极大共鸣,单日斩获 783 星,证明了开发者对 Agent 产出物质量的要求正在从“能运行”向“高品质交互与美学”跃迁。不过,由于美学标准本身具有高度主观性,如何在规范约束与多样化创意之间取得平衡仍是后续难点。
6. openai/codex-security — 官方出品的 Agent 代码安全扫描修复套件
TypeScript · 13,097 星 · 单日 +754
由 OpenAI 官方推出的安全 CLI 与 TypeScript SDK,专门用于查找、验证和自动修复代码库中的安全漏洞。随着 AI 辅助生成的代码量呈指数级增长,引入隐蔽安全隐患的风险陡增。该工具不仅能够做被动漏洞静态分析,还能借助 Codex 的理解力实现从发现到修复的闭环验证。对于正在大规模落地 AI 编码工具的技术团队而言,此类具备自动化修复能力的官方安全工具几乎已成为流水线上的标配,但也要求团队建立对自动化安全补丁的独立回归评估机制。
7. addyosmani/agent-skills — 生产级 AI 编码智能体专业技能包
JavaScript · 117,126 星 · 单日 +696
由资深前端专家 Addy Osmani 发起的项目,专注于为 AI 编码 Agent 沉淀开箱即用的生产级工程技能规范。随着各类 Agent 平台相继开放技能插件标准,模型原生能力难以覆盖复杂工程链路的缺陷愈发明显。该仓库汇集了从架构评审、复杂重构到性能优化的成熟工作流,相当于为通用 Agent 装备了资深工程师的执行脚本。其高星表现反映了工程界对于将“人类专家经验”抽象固化并注入 Agent 的急切期待,但这也对使用者自身甄别高质量技能包的能力提出了要求。
8. ggml-org/llama.cpp — 纯 C/C++ 打造的极致大模型推理引擎
C++ · 134,866 星 · 单日 +613
作为开源 AI 领域的基石级项目,llama.cpp 依然保持着单日增星 613 的强劲势头。它通过纯 C/C++ 实现并辅以深度的硬件量化优化,使得大模型能够在消费级硬件和纯 CPU/GPU 混合环境下极速运行。随着云端智能体推理开销的居高不下,越来越多的开发者选择在本地搭建轻量化私有模型来充当 Agent 的执行器或子任务分发器,llama.cpp 因此重新迎来了一波与各类 Agent Harness 联动的生态增量。其高确定性与零外部依赖特性,继续奠定着其端侧部署王者的地位。
技术趋势深度分析
第一,平台厂商与行业机构正加速争夺 Agent“规则制定权”。 从前一段时间各大厂商竞相发布参数规模更大的模型,转变为当下争夺如何约束与指导 Agent 的行为逻辑。Google Chrome 推出的 GoogleChrome/modern-web-guidance-src 与 OpenAI 发布的 openai/codex-security 均表明,头部技术力量已经意识到仅仅依靠通用预训练不足以让 Agent 编写可靠代码,必须通过结构化规范和安全 SDK 进行强干预。
第二,Agent 技能(Skills)与环境载体(Harness)正在走向模块化分工。 诸如 addyosmani/agent-skills 和 Leonxlnx/taste-skill 等项目的热度说明,开发者更倾向于把 Agent 视为一个可插拔核心,通过外挂不同的“技能卡片”来快速适配前端审美、系统重构、营销运营等多场景需求。同时,像 CopilotKit/OpenBot 这样提供完整环境沙箱的平台,正在将 Agent 的执行空间从单一文件系统扩展到包含浏览器与外部工具的自主工作台。
第三,开发流程向“人机协同治理与闭环管控”加速重构。 随着自动生成的代码量爆发,工程痛点正从“怎么写得更快”转移到“如何保证质量与控制成本”。阿里开源的 alibaba/open-code-review 与腾讯推出的 Tencent/LoopForge 均将重点放在了确定性规则结合大模型的精细化代码审查以及人工审批流上。同时,诸如 headroomlabs-ai/headroom 等上下文压缩方案的走红,也印证了团队在实际落地 Agent 时对 Token 成本和有效信息密度的严苛考量。
深度项目解析
GoogleChrome/modern-web-guidance-src:为 AI 编码设定现代 Web 底线
该项目是 Google Chrome 官方直接面向编码智能体推出的规则基准库。长期以来,大模型在生成 Web 前端代码时,普遍存在依赖过时 API、使用废弃 CSS 属性或违背无障碍设计原则的弊端。这并非因为模型推理能力不足,而是因为海量训练集中沉淀了过去二十年的历史包袱。
Chrome 团队通过梳理最新的核心 Web 指标、原生 CSS 新特性、Modern DOM API 以及响应式与无障碍标准,将其提炼为结构清晰、易于 Agent 提取解析的知识集。当开发者将此资源注入 Claude Code、Cursor 或 Codex 等智能体的上下文后,Agent 能够精准识别并规避老旧写法,直接输出符合现代浏览器标准的干净代码。
对于广大 Web 开发者而言,将该指南接入现有开发流程,不仅能显著减少 Agent 生成代码的人工重构时间,还能在团队内部统一前端基线。但需要注意的是,现代 Web 特性在部分老旧企业级运行环境(如旧版内嵌 WebView)中可能缺乏向下兼容性,实际落地时仍需结合自身的浏览器支持矩阵配置降级策略。
alibaba/open-code-review:大规模工程验证的企业级混合审查体系
在今日 Go 语言榜单中高居榜首的阿里巴巴代码审查工具,展示了企业级场景下 AI 落地审查流程的实用范式。据项目自述,该工具历经阿里巴巴海量业务场景的实战打磨,采用“确定性分析流水线 + LLM Agent”的混合架构,单日星增达到 206 颗,总星数已突破 50,560 颗。
与纯靠大模型逐行阅读的通用方案不同,该项目利用成熟的静态分析规则集对空指针异常(NPE)、线程安全风险、跨站脚本(XSS)以及 SQL 注入等硬性缺陷进行确定性截流,避免了纯大模型审查容易产生的幻觉与漏报。在精确圈定问题边界后,再由 LLM Agent 输出行级精确批注与优化建议,并全面兼容 OpenAI 和 Anthropic 等模型接口。
这种“规则保底、Agent 润色与深度推理”的混合架构,为当前苦于 AI 审查高幻觉、高成本的企业级技术负责人提供了一条清晰的实施路径。它既能守住安全与可靠性的底线,又能大幅降低人工代码评审(PR Review)的时间开销。在引入时,团队建议首先基于内置规则包跑通现有代码库,再依据业务特性渐进式放宽 Agent 的建议自主权。
本期洞察
1. 官方指导集将成为 Agent 训练与运行的标准插件:通用预训练大模型的代际知识鸿沟无法靠提示词完全弥合,类似 Chrome 团队的权威规范包将成为各类编码 Agent 运行时的基础依赖。 2. AI 代码审查正在从玩具走向确定性工程实践:以规则引擎拦截硬编码漏洞、结合 Agent 进行逻辑推导的混合架构,正在成为企业引入代码自动化治理的最优解。 3. 审美与专业领域直觉成为 Agent Skills 竞争的新维度:代码“能跑”已不再是终点,针对视觉审美、工程规范与跨端适配的高质量技能包,正在快速重构开发者的工具箱。 4. 沙箱化与行为事前审查是自主 Agent 落地的安全前提:随着 Agent 被赋予独立使用浏览器和操作系统的权限,隔离沙箱与操作可追溯机制已成为工程化必须配套的安全护栏。
数据来源:GitHub Trending 2026-11-13 · 星数为抓取时快照
榜单
- 1alibaba/open-code-review
Secure, fast, efficient, battle-tested at Alibaba's scale. Hybrid architecture code review tool: deterministic pipelines + LLM Agent, precise line-level comments, built-in multi-language ruleset (NPE, thread-safety, XSS, SQL injection), OpenAI & Anthropic compatible.
- 2GoogleChrome/modern-web-guidance-src
Keep your coding agent up to date with the latest web best practices
- 3openai/codex-security
OpenAI's Codex Security CLI and TypeScript SDK for finding, validating, and fixing security vulnerabilities. npm: https://www.npmjs.com/package/@openai/codex-security
- 4ZhuLinsen/daily_stock_analysis
LLM 驱动的多市场股票智能分析系统:多源行情、实时新闻、决策看板与自动推送,支持零成本定时运行。 LLM-powered multi-market stock analysis system with multi-source market data, real-time news, decision dashboard, automated notifications, and cost-free scheduled runs.
- 5we-promise/sure
The personal finance app for everyone (by everyone)
- 6backnotprop/plannotator
The workspace to stay in the loop and on top of agents - shape, decide, plan, annotate, review code, share with team, and send all feedback to agents with one click.
- 7louislam/uptime-kuma
A fancy self-hosted monitoring tool
- 8block/buzz
A hive mind communication platform
- 9tradermonty/claude-trading-skills
Claude Code skills for equity investors and traders — market analysis, technical charting, economic calendars, screeners, and trading strategy development.
- 10tailscale/tailcat
like netcat, but over Tailscale's data plane, without Tailscale's control plane
- 11Likely7/Veyra-NRVideo
Bring PS5 GTA6 to 120FPS with DLSS 5. An advanced Windows video player and capture enhancement tool featuring real-time preview for videos, images, and capture card sources. Combine AI-powered Super Resolution, DLSS5 image enhancement, and Frame Generation to deliver a next-generation visual experience.
- 12ChromeDevTools/chrome-devtools-mcp
Chrome DevTools for coding agents
- 13apache/iceberg
Apache Iceberg
- 14addyosmani/agent-skills
Production-grade engineering skills for AI coding agents.
- 15rustdesk/rustdesk
An open-source remote desktop application designed for self-hosting, as an alternative to TeamViewer.
- 16TheAlgorithms/Java
All Algorithms implemented in Java
- 17ggml-org/llama.cpp
LLM inference in C/C++
- 18Tencent/LoopForge
Self-hosted workflows for coding-agent delivery. Orchestrate CodeBuddy, Codex and Claude Code with scripts and human approval gates. Includes Local workflows for coding assistants.
- 19
- 20Leonxlnx/taste-skill
Agent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
- 21farion1231/cc-switch
A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode,DeepSeek Harness and more. Only official website: ccswitch.io
- 22jeecgboot/JeecgBoot
一款免费开源的 AI 低代码平台,也是 AI 时代构建复杂业务系统的开源基座!权限、多租户、工作流、报表、大屏、AI 知识库等通用能力开箱即用,AI 只需生成业务差异部分。AI Skills 一句话生成完整系统——建表、前后端代码、菜单权限、BPM 流程、表单、报表、大屏全自动。内置 AI 应用平台(聊天、知识库、流程编排、MCP),兼容主流大模型
- 23wszqkzqk/PvZ-Portable
A cross-platform reimplementation of Plants vs. Zombies by the community. Aiming to bring 100% authentic experience of GOTY edition.
- 24Tencent/WeKnora
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
- 25davila7/claude-code-templates
Open-source registry and CLI for Claude Code components: agents, commands, skills, hooks, MCPs, settings and mods
- 26affaan-m/ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
- 27tw93/Pake
Turn any website into a lightweight desktop app with one command, on macOS, Windows and Linux.
- 28ashishps1/awesome-system-design-resources
Learn System Design concepts and prepare for interviews using free resources.
- 29ggml-org/whisper.cpp
Port of OpenAI's Whisper model in C/C++
- 30ollama/ollama
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
- 31avenoxai/avenoxbeyin
Your AI second brain in one command — open-source Obsidian + Claude Code system with persistent cross-session memory. avenox.lol/beyin.md
- 32CopilotKit/OpenBot
Open-source AI coworkers that each get a computer of their own: a browser, files and tools, with every action decided before it happens and recorded after. Bring any AG-UI agent.
- 33microsoft/power-platform-skills
A plugin marketplace for GitHub Copilot and other AI agents that provides Power Platform development plugins, including reusable skills, agents, and commands for building and deploying solutions.
- 34ShayanAbbas1/dbdelve
Fast and lightweight DB Client made in Rust and GPUI.
- 35termux/termux-x11
Termux X-server add-on.
- 36
- 37Gentleman-Programming/gentle-ai
Gentle-AI configures the AI coding agents you already use: Claude Code, Cursor, OpenCode, Codex, Pi, and more. Choose persistent memory, Organic-Driven Development, curated skills, MCP servers, personas, and optional bounded review. Open source, no agent lock-in.
- 38asgeirtj/system_prompts_leaks
Leaked system prompts of ChatGPT, Claude, Gemini and more. The hidden system prompt of every popular chatbot, in full, every version since May 2025. Also Anthropic Claude Code, OpenAI Codex, Google Antigravity, xAI Grok, Meta Muse, Moonshot Kimi, Alibaba Qwen, Z.ai GLM ZCode, MiniMax, Cursor, Perplexity, DeepSeek and 30+ others. Updated daily.
- 39a2ui-project/a2ui
Open-source generative UI framework: agents stream layout data, and apps render it with native components in React, Angular, Lit, Flutter, SwiftUI and Jetpack Compose.
- 40mksglu/context-mode
Free, local plugin that keeps tool output out of your coding agent's context window. Sandbox and SQLite FTS5 search, 99% fewer bytes in context. Works with Claude Code, Codex, Pi, OpenCode, Gemini CLI, Copilot CLI and 4 more agents. Zero telemetry.
- 41zellij-org/zellij
A terminal workspace with batteries included
- 42kunal-kushwaha/DSA-Bootcamp-Java
This repository consists of the code samples, assignments, and notes for the Java data structures & algorithms + interview preparation bootcamp of WeMakeDevs.
- 43boykopovar/AnyPS5
Tool for automatic PS5 executables porting to Linux and Windows
- 44modelcontextprotocol/go-sdk
The official Go SDK for Model Context Protocol servers and clients. Maintained in collaboration with Google.
- 45anthropics/claude-plugins-official
Official, Anthropic-managed directory of high quality Claude Code Plugins.
- 46PurpleDoubleD/locally-uncensored
The all-in-one local AI studio for your desktop: chat, image and video generation and a coding agent in one free, open source app. Windows and Linux. No Docker, no terminal, no cloud required.
- 47coreyhaines31/marketingskills
Marketing skills for Claude Code and AI agents. CRO, copywriting, SEO, analytics, and growth engineering.
- 48ModernRelay/omnigraph
Object-storage native graph database with branching and typed ontology
- 49YunaiV/ruoyi-vue-pro
🔥 官方推荐 🔥 RuoYi-Vue 全新 Pro 版本,优化重构所有功能。基于 Spring Boot + MyBatis Plus + Vue & Element 实现的后台管理系统 + 微信小程序,支持 RBAC 动态权限、数据权限、SaaS 多租户、Flowable 工作流、三方登录、支付、短信、商城、CRM、ERP、MES、IM、AI 大模型、IoT 物联网等功能。你的 ⭐️ Star ⭐️,是作者生发的动力!
- 50facebookresearch/faiss
A library for efficient similarity search and clustering of dense vectors.