zhouxiaoka/autoclip

🎬 AutoClip 桌面版 v1.2.0 —— 一键安装的 Mac 客户端来了 🎬 AutoClip Desktop v1.2.0 — A one-click-install Mac app AutoClip 现在有了**真正一键安装的 macOS 客户端**：下载、双击、拖进「应用程序」即可使用，无需配置任何环境。粘贴链接就能下载视频、自动出片，整条流水线在本地完成。 AutoClip now ships a **true one-click-install macOS app**: download, double-click, drag to Applications, and you're ready — no environment setup. Paste a link to download a video and auto-generate clips, with the whole pipeline running locally. ⸻ ✨ 主要更新 / Highlights • 一键安装、零依赖：安装包内置便携 Python 运行时 + 静态 ffmpeg/ffprobe，无需预装 Python 或 ffmpeg • One-click, zero-dependency install: bundled portable Python runtime + static ffmpeg/ffprobe — no Python or ffmpeg required • 端到端本地出片：粘贴 B 站/YouTube 链接 → 下载 → 字幕 → AI 大纲/评分/标题/聚类 → 切割 → 切片与合集，全程本地运行 • End-to-end local clipping: paste a Bilibili/YouTube link → download → subtitles → AI outline/scoring/titles/clustering → cut → clips & collections, all local • 本地字幕转写（按需）：无字幕视频可在「设置 → 语音转写」一键安装 Whisper 并自选模型 • Local subtitle transcription (on demand): one-click Whisper install with selectable models for videos without subtitles • 体验改进：切片标题可手动编辑、B 站多账号管理 + 登录健康监控、合集内拖拽排序、按分类组织项目 • Usability improvements: edit clip titles by hand, Bilibili multi-account management with login health monitoring, drag-and-drop ordering within collections, organize projects by category ⸻ 📦 安装说明（macOS，Apple Silicon）/ Installation (macOS, Apple Silicon) 1. 下载下方的 AutoClip.Desktop_1.2.0_aarch64.dmg，双击后将 App 拖入「应用程序」 1. Download AutoClip.Desktop_1.2.0_aarch64.dmg below, open it, and drag the app into Applications 2. 首次打开请右键点应用 → 选「打开」（ad-hoc 签名、未做 Apple 公证，用于绕过 Gatekeeper） 2. On first launch, right-click the app → Open (ad-hoc signed, not notarized — this bypasses Gatekeeper) 3. 进入「设置」填入 AI 提供商的 API Key（通义千问 / OpenAI / Gemini 等）即可开始 3. Open Settings and enter your AI provider API key (Tongyi Qianwen / OpenAI / Gemini, etc.) to get started > 当前仅提供 Apple Silicon (M 系列) 版本；Intel / Windows / Linux 包后续提供。 > Currently Apple Silicon (M-series) only; Intel / Windows / Linux builds coming later. ⸻ ⬇️ 获取方式 / Get Started • 桌面安装包：[AutoClip.Desktop_1.2.0_aarch64.dmg](https://github.com/zhouxiaoka/autoclip/releases/download/v1.2.0/AutoClip.Desktop_1.2.0_aarch64.dmg) • Desktop installer: [AutoClip.Desktop_1.2.0_aarch64.dmg](https://github.com/zhouxiaoka/autoclip/releases/download/v1.2.0/AutoClip.Desktop_1.2.0_aarch64.dmg) • 源码 / Source: `git clone https://github.com/zhouxiaoka/autoclip.git` ⸻ 🙏 感谢一路关注与反馈的你！欢迎在 Issues 提问题、提建议。 🙏 Thanks for your continued support and feedback — questions and suggestions are welcome in Issues.

🖥️ AutoClip 桌面版 v1.1.0 —— 让 macOS 客户端真正可装、可用、能出片 🖥️ AutoClip Desktop v1.1.0 — A macOS app you can actually install, run, and clip with 这一版把整条桌面链路从头打通并端到端验证：零依赖安装、本地出片、关键稳定性修复。它是 AutoClip 从「Web 系统」迈向「桌面产品」的关键一步，后续的 v1.2.0 在此基础上加入了产品分析与更多体验改进。 This release rebuilt and verified the entire desktop pipeline end to end: zero-dependency install, local clipping, and key stability fixes. It marks AutoClip's step from a web system to a desktop product — v1.2.0 builds on it with analytics and more. > 注：v1.1.0 当时未发布带安装包的 Release（发布流程的一个权限问题，已在 v1.2.0 修复）。本说明为补录，最新可下载版本请见 [v1.2.0](https://github.com/zhouxiaoka/autoclip/releases/tag/v1.2.0)。 > Note: v1.1.0 never shipped a Release with an installer at the time (a release-pipeline permission issue, fixed in v1.2.0). This note is a backfill — for the latest downloadable build see [v1.2.0](https://github.com/zhouxiaoka/autoclip/releases/tag/v1.2.0). ⸻ 🚀 主要更新 / Highlights • 零依赖安装：内置便携 Python 运行时 + 静态 ffmpeg/ffprobe，无需预装 Python 或 ffmpeg • Zero-dependency install: bundled portable Python runtime + static ffmpeg/ffprobe — no Python or ffmpeg required • 端到端出片：粘贴 B 站/YouTube 链接 → 下载 → 字幕 → AI 大纲/时间线/评分/标题/聚类 → 切割 → 切片与合集，全程本地运行 • End-to-end clipping: link → download → subtitles → AI outline/timeline/scoring/titles/clustering → cut → clips & collections, all local • 本地字幕转写（按需）：无字幕视频可在「设置 → 语音转写」一键安装 Whisper（faster-whisper）并自选模型 • Local subtitle transcription (on demand): one-click Whisper (faster-whisper) install with selectable models ⸻ 🐛 关键修复 / Key Fixes • 修复启动黑屏、项目列表卡「加载中」、重试疯狂弹窗、处理一直卡 0% 等问题 • Fixed startup black screen, stuck "loading" project list, retry popup loop, and processing stuck at 0% • 内置 ffmpeg/ffprobe 改为静态自包含版本，换机也能正常处理视频 • Bundled ffmpeg/ffprobe are now static and self-contained, working across machines ⸻ 🙏 感谢一路的关注与反馈！/ Thanks for your continued support and feedback!

这是 Autoclip 的第一个稳定版本！ 继承了 [autoclip_mvp](https://github.com/zhouxiaoka/autoclip_mvp) 800+ Star 的验证与反馈，本仓库将作为后续开发与维护的唯一主力，标志着项目进入正式迭代阶段 🚀。 This is the first stable release of Autoclip! Building on the success of [autoclip_mvp](https://github.com/zhouxiaoka/autoclip_mvp) with 800+ stars, this repository will now serve as the mainline for future development and maintenance 🚀. ⸻ ✨ 主要更新 / Highlights • 全新架构：后端采用 FastAPI + Celery + WebSocket，前端基于 React + Zustand • New Architecture: Backend powered by FastAPI + Celery + WebSocket; Frontend built with React + Zustand • 本地字幕识别：支持 Whisper 本地推理，提升长视频转写效率 • Local Subtitle Recognition: Support for local Whisper inference, enabling efficient long-video transcription • 切片流程升级：大纲提取 → 时间点定位 → 高光评分 → 自动生成标题 • Upgraded Clipping Pipeline: Outline extraction → Timestamp detection → Highlight scoring → Auto title generation • 任务系统：支持多任务管理，实时进度推送 • Task System: Multi-task management with real-time progress updates • 稳定性：引入 Poetry 依赖管理，测试覆盖率 >95% • Stability: Dependency management with Poetry, test coverage >95% ⸻ 📦 迁移提示 / Migration Notes • 原仓库 [autoclip_mvp](https://github.com/zhouxiaoka/autoclip_mvp) 已停止维护，仅作历史保留。 • The original repository [autoclip_mvp](https://github.com/zhouxiaoka/autoclip_mvp) has been deprecated and is preserved for historical reference only. • 新用户请直接使用本仓库。 • New users should directly adopt this repository. ⸻ ⬇️ 获取方式 / Get Started • 源码下载：[Source code (zip)](https://github.com/zhouxiaoka/autoclip/releases/autoclip-v1.0.0.zip) | [Source code (tar.gz)](https://github.com/zhouxiaoka/autoclip/releases/autoclip-v1.0.0.tar.gz) • Download source: [Source code (zip)](https://github.com/zhouxiaoka/autoclip/releases/autoclip-v1.0.0.zip) | [Source code (tar.gz)](https://github.com/zhouxiaoka/autoclip/releases/autoclip-v1.0.0.tar.gz) • 或通过 Git 克隆： • Or clone the repo: git clone https://github.com/zhouxiaoka/autoclip.git ⸻ 🙏 感谢所有为 MVP 贡献 Star、反馈与讨论的用户！欢迎继续关注与参与 Autoclip 的后续开发。 🙏 Thanks to all contributors and users of the MVP for their stars, feedback, and discussions! We look forward to your continued support and contributions to Autoclip.