下面是一组“在线工具生成型 AI 提示词（Prompts）”，主题聚焦在音频处理：静音检测、响度归一化与播客剪辑流程。每条 Prompt 都要求 AI 直接产出一个可运行/可构建/可部署的在线工具：包含完整项目代码、清晰文件结构、运行命令、部署说明，并提供可执行的测试用例或 QA checklist（不少于 5 条）。
在线播客静音检测 + 一键切除工具
适合把长录音里大量停顿自动剪掉，并保留可回滚的预览与时间线。
English Prompt:
Build a web app called "Silence Cutter" that detects silence in uploaded audio and lets users preview + remove silent segments.
Requirements:
- Tech stack: Next.js (App Router) + Type + React + Tailwind.
- Audio processing: ffmpeg.wasm (client-side) OR a Node.js API route (server-side) with FFmpeg; choose one and justify.
- Features: upload audio (mp3/wav/m4a), waveform preview, adjustable silence threshold (dB) and minimum silence duration (ms), non-destructive timeline (keep original), export trimmed audio.
- Output: full project source code, file tree, exact commands to run locally, build, and deploy (Docker + optional Vercel/Render).
- Include a "segments" JSON report export (kept segments + removed segments with start/end times).
- Testing: provide at least 5 tests or a QA checklist covering edge cases (very quiet speech, clipped audio, stereo vs mono, large files, mobile Safari).
Deliverables:
1) File tree
2) All source files
3) README with run/build/deploy
4) QA checklist (>=5 items)
Chinese Explanation:
生成一个可在线上传音频并自动检测静音段的工具，支持阈值/最短静音时长调参、时间线预览与一键切除，并可导出剪辑后的音频与分段 JSON 报告。
在线响度归一化（EBU R128 / LUFS）工具
适合把不同来源的音频统一到一致的响度标准，避免忽大忽小。
English Prompt:
Create a web tool "LUFS Normalizer" that normalizes audio loudness to a target integrated LUFS.
Constraints:
- Must output a runnable project with full code, file structure, and commands.
- Provide a server-side FFmpeg pipeline (Node.js + Express or Next.js route handlers) that measures loudness (loudnorm) then applies normalization.
- UI: choose target LUFS (-16, -14, -23 presets + custom), true peak limit, and output format.
- Show a before/after loudness report table.
- Security: file size limit, safe temp cleanup, and validation.
- Deployment: Dockerfile + docker-compose example.
- Testing: include a QA checklist (>=5) including clipping prevention, very short clips, multi-channel audio, corrupt input, concurrent uploads.
Deliverables: code, README, and QA checklist.
Chinese Explanation:
做一个在线响度归一化工具：上传音频后测量 LUFS 并按目标值归一化输出，提供报告、预设与自定义参数，附完整可部署源码与测试清单。
在线音频片段标记与章节（Chapters）生成工具
适合播客/课程录音做章节点与标题，导出到常见播放器可用的格式。
English Prompt:
Build an online "Podcast Chapters" tool.
Features:
- Upload audio; play with scrubber; add chapter markers (time + + optional URL).
- Export chapters as: JSON, WebVTT, and MP4 chapter data format (if applicable).
- Provide import/export so users can edit later.
- Tech: React + Type ; backend optional.
- Include full project code, file tree, run/build commands, and deployment notes.
- Testing/QA checklist >=5 items (marker precision, timezone/locale issues, mobile playback, large audio, export validation).
Chinese Explanation:
生成一个在线章节标记工具：支持播放拖动、添加/编辑章节点，导出 JSON/WebVTT 等格式，并给出完整源码、部署与 QA 清单。
在线批量音频格式转换与参数预设工具
适合把一批音频统一转成指定码率/采样率/声道，便于分发或归档。
English Prompt:
Create a web app "Batch Audio Converter".
Requirements:
- Users can drag-and-drop multiple files.
- Provide presets (Podcast MP3 128k, Voice WAV 16k mono, AAC 256k) + custom settings.
- Show per-file progress, error handling, and final download as a zip.
- Provide a robust architecture and full source code.
- Include Docker deployment.
- QA checklist >=5 items (zip integrity, long filenames, mixed formats, memory usage, retry behavior).
Chinese Explanation:
做一个在线批量音频转换工具：支持多文件拖拽、预设与自定义参数、进度与错误提示、打包下载，并提供可部署源码与 QA 清单。
在线人声增强（语音清晰度）与噪声抑制工具
适合会议录音、人声采访的降噪与清晰度提升，强调“增强处理”而非生成内容。
English Prompt:
Build a web tool "Voice Enhancer" for existing audio.
Rules:
- This is NOT for generating audio; only processing existing recordings.
- Provide noise reduction options (spectral gate, high-pass filter, de-esser preset) using FFmpeg filters where possible.
- UI: preview original vs processed; show filter chain.
- Deliver full runnable code, file tree, commands, Docker deploy, and QA checklist >=5.
Chinese Explanation:
生成一个在线人声增强工具：对现有录音做降噪、滤波与去齿音等处理，可预览对比与导出，附完整源码与 QA 清单。
在线音频裁剪 + 淡入淡出 + 片头片尾拼接工具
适合标准化播客片头片尾，快速裁剪并输出成统一格式。
English Prompt:
Create a web app "Audio Trimmer".
Requirements:
- Upload audio; select start/end; apply fade-in/fade-out; optionally append intro/outro files.
- Use FFmpeg on server; ensure safe file handling.
- Output: project code + file tree + commands + deployment + QA checklist (>=5).
Chinese Explanation:
做一个在线音频剪辑工具：支持裁剪、淡入淡出、拼接片头片尾，输出可运行项目源码、部署与 QA 清单。
在线音频响度扫描与合规报告生成工具
适合团队统一检查音频是否满足平台响度/峰值规范，并生成可归档报告。
English Prompt:
Build "Audio Compliance Report" web tool.
Requirements:
- Upload audio; analyze loudness (integrated LUFS, loudness range, true peak) and duration.
- Produce a downloadable report (JSON + HTML/PDF optional) with pass/fail against selectable standards.
- Provide full code, file tree, commands, Docker deploy.
- QA checklist >=5 including corrupted files, huge duration, multi-channel, unicode filenames, concurrent jobs.
Chinese Explanation:
生成一个在线音频合规检查工具：上传后扫描 LUFS/峰值/时长，按标准给出通过/不通过，并导出报告与完整可部署源码。
在线音频元数据（ID3/MP4）编辑与批量写入工具
适合播客标题、作者、封面、章节信息的规范化写入与批处理。
English Prompt:
Create a web tool "Audio data Editor".
Requirements:
- Batch upload; edit common fields ( , artist, album, year, genre) and write tags.
- Show preview of current tags; support export.
- Use a safe tagging library or FFmpeg data options; document the approach.
- Deliver full project code, file tree, run/build commands, deployment instructions, and QA checklist >=5.
Chinese Explanation:
做一个在线音频元数据批量编辑工具：可查看/编辑/写入 ID3/MP4 元数据，支持批量处理与导出，附完整源码与 QA 清单。
