[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-free-ai-agent-resources-bookmark-guide-zh":3,"article-related-free-ai-agent-resources-bookmark-guide-zh":28,"series-tools-3851a445-06d0-40b6-a3b7-ae900ce52a2d":77},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":11,"views":25,"created_at":26,"published_at":27,"topic_cluster_id":11},"3851a445-06d0-40b6-a3b7-ae900ce52a2d","free-ai-agent-resources-bookmark-guide-zh","值得收藏的 AI Agent 資源庫","\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Favinash201199\u002Ffree-ai-agents-resources\" target=\"_blank\" rel=\"noopener\">Free AI Agents Resources\u003C\u002Fa> 這個 GitHub 清單，現在有 621 顆 star、76 個 fork。這數字不算爆炸，但夠說明一件事：大家真的在找 AI agent 資料，而且找得很痛苦。\u003C\u002Fp>\u003Cp>講白了，a\u003Ca href=\"\u002Fnews\u002Fsecure-local-ai-agent-openclaw-nemoclaw-zh\">gent\u003C\u002Fa> 已經不是只會 demo 的玩具。它現在卡在軟體開發流程裡。問題是，教學、Notebook、框架文件，全都散在不同地方。這個 repo 的價值，就是把這些東西收進同一個入口。\u003C\u002Fp>\u003Cp>如果你最近也在看 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fai-agents-for-beginners\" target=\"_blank\" rel=\"noopener\">Microsoft AI Agents for Beginners\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain\" target=\"_blank\" rel=\"noopener\">LangChain\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph\" target=\"_blank\" rel=\"noopener\">LangGraph\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm\" target=\"_blank\" rel=\"noopener\">OpenAI Swarm\u003C\u002Fa> 這些東西，這份清單很適合拿來當地圖。你不用再靠亂搜，然後開 \u003Ca href=\"\u002Fnews\u002F6-ways-to-get-anthropic-exposure-2026-zh\">20\u003C\u002Fa> 個分頁自我折磨。\u003C\u002Fp>\u003Ch2>這個 repo 到底收了什麼\u003C\u002Fh2>\u003Cp>這份清單不是把連結亂丟一通。它有分區塊。這點很重要。因為 AI agent 的學習路線本來就很碎，從概念、程式碼、框架，到實戰案例，順序錯了就很容易看不懂。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776906244057-umwy.png\" alt=\"值得收藏的 AI Agent 資源庫\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>它先放入門課，再放實作 Notebook，最後才是框架和社群資源。這種排法比較像真的在帶人學，而不是單純蒐集網址。對新手來說，這差很多。\u003C\u002Fp>\u003Cp>你會看到像 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents\" target=\"_blank\" rel=\"noopener\">GenAI_Agents\u003C\u002Fa> 這種 45+ Jupyter notebooks 的資源，也會看到 \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002F@MicrosoftDeveloper\" target=\"_blank\" rel=\"noopener\">Microsoft Developer\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002F@NateHerko\" target=\"_blank\" rel=\"noopener\">Nate Herk\u003C\u002Fa> 這類影片教學。這組合蠻實際的。有人靠文件學，有人靠看別人邊寫邊講學，這裡兩種路線都有。\u003C\u002Fp>\u003Cul>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fai-agents-for-beginners\" target=\"_blank\" rel=\"noopener\">Microsoft AI Agents for Beginners\u003C\u002Fa>：12 課，從基礎到工具使用\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents\" target=\"_blank\" rel=\"noopener\">GenAI_Agents\u003C\u002Fa>：45+ notebooks，可直接跑\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLlrxD0HtieHgKcRjd5-8DT9TbwdlDO-OC\" target=\"_blank\" rel=\"noopener\">Microsoft Developer AI Agents for Beginners\u003C\u002Fa>：14 支影片\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fplaylist?list=PLvQWpZ46MVvgUUUBxnqLAu-JzA-6QA1o2\" target=\"_blank\" rel=\"noopener\">Nate Herk AI Agent Tutorials\u003C\u002Fa>：70+ 支實作影片\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>為什麼它比一般整理文有用\u003C\u002Fh2>\u003Cp>很多 AI agent 彙整文最大問題，就是把理論、框架、工具混在一起。看完只會更焦慮。你知道有很多東西，但不知道先學哪個，也不知道哪個真能用。\u003C\u002Fp>\u003Cp>這個 repo 比較聰明。它把資源分成教育、框架、工具、社群幾類。這樣你可以先學概念，再挑框架，再看案例。路線清楚很多。\u003C\u002Fp>\u003Cp>它收的框架也不是隨便挑的。像 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fgeekan\u002FMetaGPT\" target=\"_blank\" rel=\"noopener\">MetaGPT\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands\" target=\"_blank\" rel=\"noopener\">OpenHands\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph\" target=\"_blank\" rel=\"noopener\">LangGraph\u003C\u002Fa>，各自對應的使用情境就不一樣。前者偏協作式角色流程，後者偏軟體工程任務，LangGraph 則強在有狀態的流程控制。\u003C\u002Fp>\u003Cp>我覺得這份清單最實用的地方，是它沒有假裝所有框架都一樣成熟。它直接把你丟進現實：有些東西適合試驗，有些東西適合做產品，有些東西只是拿來看趨勢。\u003C\u002Fp>\u003Cblockquote>“The models are the easy part. The hard part is the product.” — Andrej Karpathy, \u003Ca href=\"https:\u002F\u002Fx.com\u002Fkarpathy\" target=\"_blank\" rel=\"noopener\">X profile\u003C\u002Fa>\u003C\u002Fblockquote>\u003Cp>這句話很貼切。模型現在很多，API 也不缺。真正難的是把 agent 做成能用的軟體。這份 repo 的價值，就是幫你少走一點冤枉路。\u003C\u002Fp>\u003Ch2>主流框架怎麼比\u003C\u002Fh2>\u003Cp>如果你真的要做 agent，框架選擇很重要。不是每個工具都適合一樣的任務。有人想做自動化流程，有人想做 coding assistant，有人只想快速驗證想法。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776906235157-g7ii.png\" alt=\"值得收藏的 AI Agent 資源庫\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>這份清單裡的幾個代表，剛好能看出市場分工。\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm\" target=\"_blank\" rel=\"noopener\">OpenAI Swarm\u003C\u002Fa> 偏輕量協調；\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain\" target=\"_blank\" rel=\"noopener\">LangChain\u003C\u002Fa> 還是通用型選項；\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph\" target=\"_blank\" rel=\"noopener\">LangGraph\u003C\u002Fa> 更適合有分支、有狀態的流程。\u003C\u002Fp>\u003Cp>如果你做的是軟體工程任務，\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands\" target=\"_blank\" rel=\"noopener\">OpenHands\u003C\u002Fa> 很值得盯。它不是拿來賣概念的。它就是想讓 agent 幫你寫、改、查、修程式。這種方向比較接近真實團隊需求。\u003C\u002Fp>\u003Cul>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fai-agents-for-beginners\" target=\"_blank\" rel=\"noopener\">Microsoft AI Agents for Beginners\u003C\u002Fa>：12 課，適合新手\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents\" target=\"_blank\" rel=\"noopener\">GenAI_Agents\u003C\u002Fa>：45+ notebooks，適合邊看邊跑\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph\" target=\"_blank\" rel=\"noopener\">LangGraph\u003C\u002Fa>：適合狀態機與分支流程\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands\" target=\"_blank\" rel=\"noopener\">OpenHands\u003C\u002Fa>：適合 coding 與 debugging\u003C\u002Fli>\u003Cli>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Freworkd\u002FAgentGPT\" target=\"_blank\" rel=\"noopener\">AgentGPT\u003C\u002Fa>：偏視覺化、無程式碼實驗\u003C\u002Fli>\u003C\u002Ful>\u003Cp>如果你只想快速玩看看，\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Freworkd\u002FAgentGPT\" target=\"_blank\" rel=\"noopener\">AgentGPT\u003C\u002Fa> 這種視覺化工具比較友善。你要做比較像產品的東西，就會更常碰到 LangGraph 或 \u003Ca href=\"\u002Fnews\u002Fflorida-criminal-probe-openai-chatgpt-zh\">Open\u003C\u002Fa>Hands。這差別很現實，也很殘酷。\u003C\u002Fp>\u003Ch2>誰該先收藏這份清單\u003C\u002Fh2>\u003Cp>如果你是新手，先從入門課開始。真的不要一開始就跳進複雜框架。你會很快卡在 token、tool calling、memory、planning 這些詞上，然後開始懷疑人生。\u003C\u002Fp>\u003Cp>如果你已經在做 LLM 產品，那你可以直接看框架和除錯資源。像 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fonestardao\u002FWFGY\" target=\"_blank\" rel=\"noopener\">WFGY\u003C\u002Fa> 這種工具，就比較接近實戰。你需要的是定位問題，不是再看一篇空泛文章。\u003C\u002Fp>\u003Cp>如果你是團隊 lead，這份 repo 也能當內部教材。把它當共用閱讀清單，至少可以少掉那種「到底要看哪篇」的會議。說真的，這種時間很值錢。\u003C\u002Fp>\u003Cp>更重要的是，agent 已經不是單純 prompt engineering。現在比的是工具調用、記憶管理、流程控制、錯誤恢復。這些東西都偏工程。會做的人，和只會聊模型的人，差距會越來越明顯。\u003C\u002Fp>\u003Ch2>這波資源整理背後的產業脈絡\u003C\u002Fh2>\u003Cp>AI agent 這一波熱度，來自兩個方向。第一個是模型能力變強。第二個是開發者開始真的想把它接進工作流。這時候，光懂聊天不夠。你得懂 API、資料流、狀態、重試機制，還有怎麼把失敗處理好。\u003C\u002Fp>\u003Cp>這也是為什麼 GitHub 上的資源整理會變重要。因為官方文件通常只講單一產品。社群整理則會把不同框架放在一起。對開發者來說，這種比較才有用。\u003C\u002Fp>\u003Cp>台灣團隊如果在看 AI agent，我會建議先想清楚三件事：你要自動化什麼、你能接受多少錯誤、你要不要可觀測性。這三題沒想清楚，後面很容易做出一個看起來很炫，實際上很難維護的東西。\u003C\u002Fp>\u003Cp>從這角度看，這份清單不是只給學生看。它也適合做產品規劃前的掃描。你可以先看目前有哪些框架，再決定自己要不要做 orchestration layer，或直接接現成工具。\u003C\u002Fp>\u003Ch2>結論：先學一條線，再做一個小專案\u003C\u002Fh2>\u003Cp>我會建議你先選一條路。入門課選一個，框架選一個，除錯工具選一個。不要一次全收。那樣只會變成收藏夾收藏家。\u003C\u002Fp>\u003Cp>如果你本月真的想動手，我的建議很簡單：先跑 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fai-agents-for-beginners\" target=\"_blank\" rel=\"noopener\">Microsoft AI Agents for Beginners\u003C\u002Fa>，再挑 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph\" target=\"_blank\" rel=\"noopener\">LangGraph\u003C\u002Fa> 或 \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands\" target=\"_blank\" rel=\"noopener\">OpenHands\u003C\u002Fa> 做一個小專案。像是文件問答、任務分派、或自動修 bug，都很適合。\u003C\u002Fp>\u003Cp>這份 GitHub 清單最實際的用途，不是讓你看完覺得自己懂了。是讓你少繞路。你如果真想進 AI agent，現在就該開始挑一條能跑的路線，而不是繼續滑影片收藏連結。\u003C\u002Fp>","一個 GitHub 彙整頁收進 Microsoft、LangChain、OpenAI 等 AI agent 教學與框架。適合想從入門課、Notebook 到實作框架，一次整理學習路線的開發者。","github.com","https:\u002F\u002Fgithub.com\u002Favinash201199\u002Ffree-ai-agents-resources",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1776906244057-umwy.png","tools","zh","8bce028d-4ea9-493b-8a9f-df5a1b373838",[17,18,19,20,21,22,23,24],"AI agent","GitHub 資源","LangChain","LangGraph","OpenAI Swarm","OpenHands","Microsoft AI Agents for Beginners","GenAI_Agents",8,"2026-04-23T01:03:39.387934+00:00","2026-04-23T01:03:39.158+00:00",{"tags":29,"relatedLang":36,"relatedPosts":40},[30,32,34],{"name":20,"slug":31},"langgraph",{"name":19,"slug":33},"langchain",{"name":17,"slug":35},"ai-agent",{"id":15,"slug":37,"title":38,"language":39},"free-ai-agent-resources-bookmark-guide-en","Free AI Agent Resources Worth Bookmarking","en",[41,47,53,59,65,71],{"id":42,"slug":43,"title":44,"cover_image":45,"image_url":45,"created_at":46,"category":13},"6d635736-4581-484c-aa7f-9c221b54215b","pixelrag-screenshots-retrievable-context-zh","PixelRAG 把截圖變可檢索上下文","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782759807854-613a.png","2026-06-29T19:02:59.448799+00:00",{"id":48,"slug":49,"title":50,"cover_image":51,"image_url":51,"created_at":52,"category":13},"cdfe98b4-cfc9-4266-ba45-4b892e8214b0","codex-deepseek-v4-pro-moark-setup-zh","Codex 接入 DeepSeek-V4-Pro，三步可用","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782738175563-wcfd.png","2026-06-29T13:02:24.814293+00:00",{"id":54,"slug":55,"title":56,"cover_image":57,"image_url":57,"created_at":58,"category":13},"37ef32c4-b40c-4d2e-822a-d41678837de1","devin-ai-alternatives-real-workflows-zh","Devin 替代工具先看工作流","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782732810886-up7s.png","2026-06-29T11:32:58.321746+00:00",{"id":60,"slug":61,"title":62,"cover_image":63,"image_url":63,"created_at":64,"category":13},"d98bcfa9-686a-4465-8371-ba06d04b9799","claude-code-turns-agent-setup-into-terminal-work-zh","Claude Code 讓代理設定變終端工作","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782731910840-j284.png","2026-06-29T11:18:01.715271+00:00",{"id":66,"slug":67,"title":68,"cover_image":69,"image_url":69,"created_at":70,"category":13},"054589c7-78d8-4835-a5e4-4413a6146970","best-ai-coding-agent-2026-ranked-benchmarks-zh","2026 最佳 AI 寫碼代理排名","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782730985842-v7mt.png","2026-06-29T11:02:38.646028+00:00",{"id":72,"slug":73,"title":74,"cover_image":75,"image_url":75,"created_at":76,"category":13},"f1ea1e37-c8c4-4b45-828a-b120c781bd6f","openclaw-bailian-qwen37-max-config-template-zh","OpenClaw 接百炼 Qwen3.7-Max 模板","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782715688788-y9u4.png","2026-06-29T06:47:44.371703+00:00",[78,83,88,93,98,103,108,113,118,123],{"id":79,"slug":80,"title":81,"created_at":82},"855cd52f-6fab-46cc-a7c1-42195e8a0de4","surepath-real-time-mcp-policy-controls-zh","SurePath 推出即時 MCP 政策控管","2026-03-26T07:57:40.77233+00:00",{"id":84,"slug":85,"title":86,"created_at":87},"9b19ab54-edef-4dbd-9ce4-a51e4bae4ebb","mcp-in-2026-the-ai-tool-layer-teams-use-zh","2026 年 MCP：團隊真的在用的 AI 工具層","2026-03-26T08:01:46.589694+00:00",{"id":89,"slug":90,"title":91,"created_at":92},"af9c46c3-7a28-410b-9f04-32b3de30a68c","prompting-in-2026-what-actually-works-zh","2026 提示工程，真正有用的是什麼","2026-03-26T08:08:12.453028+00:00",{"id":94,"slug":95,"title":96,"created_at":97},"05553086-6ed0-4758-81fd-6cab24b575e0","garry-tan-open-sources-claude-code-toolkit-zh","Garry Tan 開源 Claude Code 工具包","2026-03-26T08:26:20.068737+00:00",{"id":99,"slug":100,"title":101,"created_at":102},"042a73a2-18a2-433d-9e8f-9802b9559aac","github-ai-projects-to-watch-in-2026-zh","2026 必看 20 個 GitHub AI 專案","2026-03-26T08:28:09.619964+00:00",{"id":104,"slug":105,"title":106,"created_at":107},"a5f94120-ac0d-4483-9a8b-63590071ac6a","claude-code-vs-cursor-2026-zh","Claude Code 與 Cursor 深度對比：202…","2026-03-26T13:27:14.279193+00:00",{"id":109,"slug":110,"title":111,"created_at":112},"0975afa1-e0c7-4130-a20d-d890eaed995e","practical-github-guide-learning-ml-2026-zh","2026 機器學習入門 GitHub 實用指南","2026-03-27T01:16:49.712576+00:00",{"id":114,"slug":115,"title":116,"created_at":117},"bfdb467a-290f-4a80-b3a9-6f081afb6dff","aiml-2026-student-ai-ml-lab-repo-review-zh","AIML-2026：像課綱的學生實驗 Repo","2026-03-27T01:21:51.467798+00:00",{"id":119,"slug":120,"title":121,"created_at":122},"80cabc3e-09fc-4ff5-8f07-b8d68f5ae545","ai-trending-github-repos-and-research-feeds-zh","AI Trending：把 AI 資源收成一張表","2026-03-27T01:31:35.262183+00:00",{"id":124,"slug":125,"title":126,"created_at":127},"3ce6e6e2-bac5-463e-9f8d-45caabcc61f7","awesome-ai-for-science-research-tools-map-zh","AI 科研工具清單，開始像地圖了","2026-03-27T01:46:50.521945+00:00"]