<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Ziyu 的 AI 影片工作室筆記</title><description>AI 影片工作室與軟體工程的實作筆記——全本地生成、代理協作、雙平台發佈</description><link>https://ziyu4huang.github.io/blogs/</link><language>zh-TW</language><item><title>我的 AI 影片頻道總覽:四個系列,全部在一台 MacBook 上生成</title><link>https://ziyu4huang.github.io/blogs/posts/2026-10-04-ai-video-channel-guide/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-10-04-ai-video-channel-guide/</guid><description>本站對應的影片頻道總索引——《跟著致敬片學 AI 分鏡》教學系列、三個致敬/復刻系列,每一集的 B站 與 YouTube 連結都在這,持續更新。</description><pubDate>Sun, 04 Oct 2026 06:10:00 GMT</pubDate><category>ai-video</category><category>channel</category></item><item><title>h3c 客製全紀錄:從 antirez 的 h3.c 到量產 Metal 影片引擎</title><link>https://ziyu4huang.github.io/blogs/posts/2026-10-04-h3c-engine-custom-build/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-10-04-h3c-engine-custom-build/</guid><description>C11 + Metal 直讀 safetensors、三顆引擎樹的版本紀律、七項 fork 客訂——控制 tokens、音訊 sigma schedule、pruned AdaLN、runtime LoRA 檢查——加上互動佇列,我們把 MiniMax H3 鍛成一顆在 MacBook 上量產有聲影片的本機引擎。</description><pubDate>Sun, 04 Oct 2026 04:00:00 GMT</pubDate><category>ai-video</category><category>h3c</category><category>metal</category><category>apple-silicon</category></item><item><title>生產宣言:純本地、零雲端的指揮式 AI 影片工作室</title><link>https://ziyu4huang.github.io/blogs/posts/2026-10-04-local-video-studio-philosophy/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-10-04-local-video-studio-philosophy/</guid><description>在一台 MacBook Pro M5 Max 上,人只下命令,AI 自主完成規劃、prompt、渲染、QC、修復到發佈——這是我經營本地 AI 影片工作室的三條鐵則與配套原則。</description><pubDate>Sun, 04 Oct 2026 00:00:00 GMT</pubDate><category>ai-video</category><category>local-first</category><category>agent-workflows</category></item><item><title>MiniMax H3 品質基準測試:在 Apple Silicon 上找到影片生成的甜點</title><link>https://ziyu4huang.github.io/blogs/posts/2026-10-04-minimax-h3-quality-benchmark/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-10-04-minimax-h3-quality-benchmark/</guid><description>從 20 步到 30 步、turbo LoRA 到完整 66GB transformer、608×352 到 720p——一份在 M5 Max 上實測的 MiniMax H3 影片品質階梯,附七個 Phase C 實驗的判決紀錄。</description><pubDate>Sun, 04 Oct 2026 00:00:00 GMT</pubDate><category>ai-video</category><category>benchmark</category><category>apple-silicon</category></item><item><title>Fubon Neo API: A Developer&apos;s Guide to Taiwan Stock Market Data</title><link>https://ziyu4huang.github.io/blogs/posts/2026-04-09-fubon-neo-api-taiwan-stock/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-04-09-fubon-neo-api-taiwan-stock/</guid><description>A developer tour of the Fubon Neo API for Taiwan market data: auth, quotes, rate limits, and TypeScript patterns for turning raw ticks into usable signals.</description><pubDate>Thu, 09 Apr 2026 02:00:00 GMT</pubDate><category>API</category><category>TypeScript</category><category>Taiwan Stock</category><category>Fubon</category><category>Bun</category><category>Financial Data</category></item><item><title>Building a Taiwan Stock Market Explainer Video with Remotion + Claude Code</title><link>https://ziyu4huang.github.io/blogs/posts/2026-04-08-remotion-claude-code-taiwan-stock-video/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2026-04-08-remotion-claude-code-taiwan-stock-video/</guid><description>Building a Taiwan stock explainer video end-to-end with Remotion and Claude Code: data pipeline, scene composition, narration, and the render tricks that made it watchable.</description><pubDate>Wed, 08 Apr 2026 02:00:00 GMT</pubDate><category>Remotion</category><category>Claude Code</category><category>TypeScript</category><category>React</category><category>Video Generation</category><category>AI</category><category>Bun</category></item><item><title>Nano Banana: Google&apos;s AI Revolution in Image Generation and Editing</title><link>https://ziyu4huang.github.io/blogs/posts/2025-10-06-nano-banana-ai-revolution/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2025-10-06-nano-banana-ai-revolution/</guid><description>Hands-on with Google Nano Banana: what it generates well, how its edit mode compares to plain re-prompting, and a gallery of generation and edit examples.</description><pubDate>Mon, 06 Oct 2025 12:59:10 GMT</pubDate><category>AI</category><category>Google</category><category>Image Generation</category><category>Gemini</category><category>Nano Banana</category></item><item><title>Why We Need Claude Code SDK: Pros, Cons, and the Future of AI Development Tools</title><link>https://ziyu4huang.github.io/blogs/posts/2025-10-06-why-we-need-claude-code-sdk/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2025-10-06-why-we-need-claude-code-sdk/</guid><description>Where the Claude Code SDK shines, where it fights you, and how to choose between SDK, CLI and raw API when automating agent workflows.</description><pubDate>Mon, 06 Oct 2025 00:30:00 GMT</pubDate><category>Claude Code</category><category>SDK</category><category>AI development</category><category>programming</category><category>developer tools</category><category>automation</category></item><item><title>Configure Z.ai with Claude Code: Complete Guide to GLM-4.6 Integration</title><link>https://ziyu4huang.github.io/blogs/posts/2025-10-04-configure-zai-claude-code/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2025-10-04-configure-zai-claude-code/</guid><description>Step-by-step guide to wiring Z.ai GLM-4.6 into Claude Code: base URL, auth token, model mapping, and the pitfalls (context limits, MCP output caps) hit along the way.</description><pubDate>Sat, 04 Oct 2025 03:00:00 GMT</pubDate><category>Z.ai</category><category>Claude Code</category><category>GLM-4.6</category><category>AI development</category><category>API integration</category><category>coding</category></item><item><title>Introducing BMAD-METHOD: A Universal AI Agent Framework</title><link>https://ziyu4huang.github.io/blogs/posts/2025-10-04-introducing-bmad-method/</link><guid isPermaLink="true">https://ziyu4huang.github.io/blogs/posts/2025-10-04-introducing-bmad-method/</guid><description>What the BMAD-METHOD multi-agent framework actually does, how its analyst-architect-dev-QA role chain maps onto real builds, and when it beats a single well-prompted session.</description><pubDate>Sat, 04 Oct 2025 02:32:00 GMT</pubDate><category>AI</category><category>BMAD</category><category>agents</category><category>development</category><category>framework</category></item></channel></rss>