近 30 天新建、star 最多的 AI 开源项目 Top 30（数据来自 GitHub 官方 API，每日快照）。
今日导读
本周 AI 开源圈最值得关注的动向，是编码智能体（coding agent）基础设施的快速成型。xAI 开源的 grok-build 以两万三千余星强势领跑，作为一套带全屏 TUI 的编码 agent 执行框架，它的出现标志着大模型厂商开始亲自下场定义 agent 运行时的交互规范，而不再只是提供 API。与此同时，unicity-aos 的 AOS Community Edition 试图在更底层构建"智能体操作系统"，思路更激进——如果这个方向成立，未来的 agent 可能不是跑在应用层，而是跑在一套专门的运行时之上。 另一个值得关注的侧面是内容质量工具的崛起。no-ai-slop 用规则集清除写作中的 AI 腔调，三千多星的热度说明开发者对 AI 生成内容的审美疲劳已经到了需要工具介入的程度——这本身就是一个信号。整体来看，这批项目的共同气质是"让 AI 真正能干活"，从 agent 框架到红队测试平台，基础设施层的竞争正在加速。
| # | 项目 | Star | 语言 | 创建 |
|---|---|---|---|---|
| 1 | xai-org/grok-build SpaceXAI's coding agent harness and TUI. Fullscreen, mouse interactive, extensible. | 23,316 | Rust | 2026-07-14 |
| 2 | unicity-aos/aos-ce AOS Community Edition: the open agent operating system. | 7,746 | Rust | 2026-07-12 |
| 3 | elder-plinius/T3MP3ST autonomous red teaming platform; multi-agent offensive-security meta-harness | 5,281 | TypeScript | 2026-07-02 |
| 4 | MDX-Tom/gpt-5.6-instruct A Codex CLI jailbreak prompt and test pack for gpt-5.6-sol. 针对 gpt-5.6 系列的 Codex CLI 破甲提示词与测试包。 | 3,599 | Python | 2026-07-11 |
| 5 | petergyang/no-ai-slop Removes 20+ patterns of AI slop from any piece of writing. | 3,316 | Python | 2026-07-07 |
| 6 | synthetic-sciences/openscience The open-source AI workbench for scientific research | 2,920 | TypeScript | 2026-07-03 |
| 7 | isjiamu/gzh-design-skill 把 Markdown 一键排成可直接粘进公众号编辑器的精致 HTML —— 6 套精选主题 + 主题生成器 + 双关卡校验。An AI-agent skill that turns Markdown into paste-ready WeChat article HTML. | 2,615 | HTML | 2026-07-01 |
| 8 | Vincentwei1021/video-shotcraft AI video skill for Claude Code & Codex — cinematic product videos with Remotion: 106 shot recipe cards, 161 motion previews, a production-ready template | 2,578 | TypeScript | 2026-07-19 |
| 9 | lopopolo/harness-engineering 🐎 Ryan Lopopolo’s anthology, field guide, and agent context bundle for harness engineering | 2,397 | Python | 2026-07-18 |
| 10 | slvDev/esp32-ai | 2,130 | Python | 2026-07-23 |
| 11 | Sahir619/fable-method The Fable Workflow: how Claude Fable 5 worked, distilled into skills any model can run, with the eval that keeps it honest. Think / act / prove. | 1,962 | Python | 2026-07-06 |
| 12 | HUANGCHIHHUNGLeo/claude-real-video Let Claude (or any LLM) actually watch a video — scene-aware, deduplicated frames + transcript, from a URL or local file. Runs locally, MIT. | 1,892 | Python | 2026-06-30 |
| 13 | penecho/penecho Think with AI beyond the chat box. A shared canvas for handwriting, equations, diagrams, and spatial reasoning. | 1,775 | JavaScript | 2026-07-14 |
| 14 | jamesob/local-llm Everything I know about running LLMs locally | 1,672 | Shell | 2026-07-03 |
| 15 | MIgHTy-alIeN/MEV-Arbitrage-Bot An arbitrage bot is a smart contract connected to an external automation script that controls its operation. | 1,611 | Solidity | 2026-07-17 |
| 16 | tandpfun/wardrobe Your clothes, extracted and organized with gpt-image. | 1,597 | JavaScript | 2026-07-16 |
| 17 | buchidonggua/dg-ai-notes | 1,540 | MDX | 2026-07-05 |
| 18 | kvcache-ai/AgentENV AgentENV (AENV) is a distributed platform for running agent environments at scale. | 1,532 | Rust | 2026-07-23 |
| 19 | AlephAITech/WorkBuddyGuide A practical, open-source guide to mastering WorkBuddy through real-world workflows.开源的 WorkBuddy 实战蓝皮书：教程、真实工作流、Skills、MCP、自动化与多智能体实践。 | 1,487 | Python | 2026-07-10 |
| 20 | jakubkrehel/skills A collection of agent skills that help with various parts of building a great interface. From animation and UI polish to accessibility and product writing. | 1,451 | — | 2026-07-10 |
| 21 | Jakubantalik/thinking-orbs Dotted thought-orb loading indicators for AI & agent UIs — six tuned states, two sizes, auto dark/light | 1,228 | TypeScript | 2026-07-21 |
| 22 | v-modal/vmodal_sdk_flutter V- Modal AI: Search anything anywhere SDK Flutter | 1,207 | Dart | 2026-07-16 |
| 23 | nethical6/conversation-steganography Use LLMs to hide messages inside normal looking conversations | 1,152 | Go | 2026-07-17 |
| 24 | asz798838958/freeAgentIdentity 多线程全自动注册free 绕过接码使用codex | 1,142 | Python | 2026-07-02 |
| 25 | simonlin1212/Vibe-Research Vibe-Research: Your Personal Trading Research Agent · A股/美股/港股 的个人投研 Agent：每日复盘、资讯雷达、个股数据、板块中心、我的持仓、研究记录。Vibe-Research 把数据和功能配齐，由你自己的 AI 驱动投资研究。 | 1,054 | TypeScript | 2026-07-05 |
| 26 | makecindy/cindy Consider it done. The open-source AI agent that works out of the box · 想到，就能做到。开源、开箱即用的 AI Agent。 | 1,044 | TypeScript | 2026-07-22 |
| 27 | aipoch/open-science An open-source, model-agnostic AI workbench for scientific discovery. | 1,037 | TypeScript | 2026-07-03 |
| 28 | ai4s-research/open-science Open Science Desktop — local-first, model-agnostic AI research workbench for macOS, Windows & Linux. Open-source Claude Science desktop alternative built on Tauri + MCP + agent skills. | 1,021 | TypeScript | 2026-07-03 |
| 29 | SmileLikeYe/agent-chief Attention is your scarcest resource. Chief is the local-first layer that guards it — turning every agent, alert, and feed into one honest call: interrupt, or not. | 1,016 | Python | 2026-07-04 |
| 30 | Blaizzy/nativ Local AI, native to your Mac. Chat, serve, monitor, and connect MLX models from one macOS app. | 1,011 | Swift | 2026-07-20 |
历史日报
数据来源 GitHub Search API；「近 30 天新建 + star 排序」反映新项目热度，与 GitHub Trending 口径略有不同。项目描述为原文，导读为本站编辑内容。