本篇只聚焦【视频处理】类在线工具生成型提示词（Prompts）。每条 Prompt 都要求 AI 直接产出：完整项目代码、清晰文件结构、可复制运行命令、部署步骤，以及不少于 5 条测试用例或 QA checklist，用来确保工具可运行、可构建、可上线。
在线视频编码参数体检（ffprobe 报告 + 合规建议）工具
上传或粘贴视频链接，自动解析容器/编码信息，输出可读报告与平台兼容性检查项。
英文 Prompt:
Build a production-ready web app called "Video Encode Inspector".
Goal:
- Let users upload a video file (MP4/MOV/MKV) or provide a direct URL.
- Run ffprobe on the server to extract container + stream data.
- Render a human-friendly report (codec, profile, level, bitrate, GOP/keyframe interval, resolution, fps, color range, audio codec/channels/sample rate, sub s).
- Provide compatibility checks presets (Web/HLS, iOS Safari, Android, Smart TV) with pass/warn/fail rules.
Tech constraints:
- Use Next.js (App Router) + Type .
- Backend runs in Node.js with a queue (BullMQ + Redis) to process large videos.
- Store uploads in local disk for dev, S3-compatible for prod.
- Security: validate MIME/type, size limit, virus-scan hook placeholder, signed URL downloads.
Deliverables (MUST output all):
1) Complete file tree.
2) Full source code for all files.
3) Dev run commands + env examples.
4) Dockerfile + docker-compose (Next.js + Redis).
5) Deployment guide (VPS + reverse proxy + env + storage).
6) Tests: at least 5 test cases (unit/integration) + a QA checklist.
7) Provide sample ffprobe JSON parsing and rule engine implementation.
UI requirements:
- Upload area, URL input, job progress, report page, export report as JSON and Markdown.
中文释义:
生成一个可上线的“视频参数体检”在线工具：能用 ffprobe 抽取视频元信息，按预设规则给出兼容性建议，并提供报告导出、队列与存储等生产级细节。
在线转码预设生成器（平台模板 + 一键生成 ffmpeg 命令）
面向不同平台输出统一的转码参数，并生成可复制命令与可下载的预设文件。
英文 Prompt:
Create a web tool "FFmpeg Preset Builder".
Features:
- Choose target platform preset: YouTube, TikTok, WeChat, HLS streaming, Web MP4.
- Users set constraints: max bitrate, max width/height, fps, audio bitrate, two-pass toggle.
- Generate:
(a) ffmpeg command lines
(b) a JSON preset file
(c) a Bash/PowerShell for batch processing a folder
- Include a validate preset step that warns about incompatible combinations.
Implementation:
- React + Type frontend.
- Node.js API for preset validation and generation.
- No actual video processing required for this tool, but provide an optional plugin interface.
Deliverables:
- Full project code + file tree.
- Run commands.
- Deployment instructions.
- Tests (>=5) for preset validation and generation.
- QA checklist.
中文释义:
做一个“转码预设生成器”在线工具：让用户按平台模板选择参数，自动生成 ffmpeg 命令与批处理脚本，并提供规则校验与测试用例。
在线批量转码工作台（任务队列 + 进度 + 失败重试）
把一堆视频扔进去，按同一预设批量转码，实时展示进度与日志，失败可重试。
英文 Prompt:
Build "Batch Transcode Workbench" as a web app.
Core:
- Upload multiple videos; choose output container (mp4) and preset (H.264/AAC).
- Server-side processing with ffmpeg and a job queue.
- Show per-job progress (parse ffmpeg stderr), ETA, logs, and downloadable outputs.
- Retry failed jobs; cancel running jobs.
Tech:
- Next.js + Type .
- Worker process for ffmpeg jobs.
- Storage: local dev + S3 prod.
- Observability: structured logs, basic metrics endpoint.
Deliverables:
- Complete source code + file tree.
- Docker compose for web + worker + redis.
- Deployment guide.
- Tests: queue, progress parser, cancellation, storage.
- QA checklist.
Important:
- Do not generate images; focus on video processing only.
中文释义:
生成一个带队列与进度条的“在线批量转码工作台”：包含 worker、日志、失败重试、取消任务、存储与部署细节，并提供可验证的测试与 QA。
在线视频封装/码流检查器（容器结构、时长漂移、音画不同步风险）
分析时间戳、帧率模式、B 帧/参考帧等，帮助排查播放卡顿与音画不同步。
英文 Prompt:
Create "Video Container Analyzer" web tool.
Functionality:
- Upload a video.
- Run ffprobe + additional ffmpeg analysis commands to detect:
- VFR vs CFR
- missing PTS/DTS
- non-monotonic DTS
- negative timestamps
- audio/video duration mismatch
- keyframe interval statistics
- Produce a diagnostic report with suggested fixes (example ffmpeg remux commands).
Requirements:
- Provide complete project code (Type ), file tree, run/deploy commands.
- Include at least 5 automated tests + QA checklist.
- Provide sample videos in tests as fixtures via tiny synthetic generation commands (ffmpeg) in test setup.
中文释义:
生成一个“视频封装与码流诊断”在线工具：能检测常见时间戳/封装问题并给出修复命令示例，同时提供测试夹具与可复现的验证流程。
在线视频音频规范化与响度检测（LUFS/Peak）报告工具
对视频中的音频做响度检测，输出适配不同平台的建议值与处理命令。
英文 Prompt:
Build "Video Loudness Report" web app.
What it does:
- Upload video.
- Extract audio and run loudness analysis (EBU R128 / LUFS, true peak).
- Output a report and recommended normalization command examples.
Stack:
- Node.js + ffmpeg on server.
- Simple React UI.
Deliverables:
- Full code + file tree.
- CLI commands for local dev.
- Docker deployment.
- Tests (>=5) for parsing analysis output and generating recommendations.
- QA checklist.
中文释义:
生成一个“视频响度检测报告”工具：专注于从视频提取音频并计算 LUFS/Peak，输出报告与规范化建议，配套测试与部署说明。
在线视频裁剪与片段导出工具（无重新编码/可选重编码）
支持按时间码裁剪、导出片段，并给出“无损裁剪”与“重编码裁剪”的差异说明。
英文 Prompt:
Create "Online Video Cutter".
Features:
- Upload a video.
- Set start/end time; preview thumbnails.
- Export clip in two modes:
1) fast cut (stream copy) when possible
2) re-encode cut (H.264/AAC) when stream copy is not safe
- Provide clear warnings about keyframes and accuracy.
Deliverables:
- Full project code + file tree.
- Commands, Docker, deployment guide.
- Tests: time parsing, boundary conditions, export modes, error handling.
- QA checklist.
中文释义:
生成一个“在线视频裁剪导出”在线工具：能给出无重编码与重编码两种导出策略，提示关键帧风险，并提供完整源码、命令、部署与测试。
在线视频字幕轨道提取与封装校验工具（不生成图片）
提取内封字幕轨道、校验编码与时间轴，导出为 srt/vtt 等格式，并提供封装检查。
英文 Prompt:
Build "Sub Track Extractor" web tool.
Must-have:
- Upload video.
- List sub streams (language, codec, default/forced flags).
- Extract to SRT/VTT when possible.
- Validate timing issues (overlaps, negative timestamps, huge gaps) and generate a validation report.
Constraints:
- Do NOT generate images; only parse/extract/validate sub text.
Deliverables:
- Complete Type project code + file tree.
- Run + deploy commands.
- Tests (>=5) for sub parsing/validation.
- QA checklist.
中文释义:
生成一个“字幕轨道提取与校验”工具：只做字幕流的解析、提取与时间轴校验，不涉及任何出图/配图，并给出可上线的完整实现与测试。
在线视频转码结果对比工具（参数/码率/帧率差异清单）
上传“原始视频 + 转码后视频”，生成差异报告，帮助验证预设是否达到目标。
英文 Prompt:
Create a web app "Transcode Diff".
Workflow:
- Upload two videos: source and output.
- Run ffprobe on both.
- Produce a diff report: resolution, fps, bitrate, codec, audio properties, duration drift, keyframe interval stats.
- Export as Markdown.
Deliverables:
- Full code + file tree.
- Dev + deployment commands.
- Tests (>=5) for diff engine and report generation.
- QA checklist.
中文释义:
生成一个“转码前后差异对比”在线工具：把两段视频的 ffprobe 信息做结构化对比，输出差异清单与报告导出，并有足够测试验证逻辑正确。
在线视频发布前自检清单生成器（编码规则可配置）
把平台要求写成可配置规则，生成发布前 QA 检查项与自动化校验建议。
英文 Prompt:
Build "Video Preflight Checklist Generator".
Goal:
- Provide a rule editor (YAML/JSON) for video delivery requirements (codec, max bitrate, resolution, audio codec, loudness range, captions required).
- Given a video (uploaded) generate:
- pass/warn/fail results
- a downloadable QA checklist for manual review
- recommended ffmpeg commands to fix common issues
Tech:
- Next.js + Type .
- Rule engine module with tests.
Deliverables:
- Full source code + file tree.
- Run commands + Docker.
- Deployment guide.
- Tests (>=5) for rule parsing and evaluation.
- QA checklist.
中文释义:
生成一个“视频发布前自检”在线工具：把编码要求配置成规则，结合视频体检结果输出 QA 清单与修复建议，包含可运行代码、部署与测试。
