手工川 LogoLovstudio.AI
Lovstudio.AI Skill / Video Creation·VIP · INCLUDED IN MEMBERSHIP

Group Interview Short

Turn a phone recording of a group dinner or salon into a vertical film where every guest appears, plus a solo short per guest, honoring each person's privacy requests

OVERVIEW

What you can do with it

群访短片:把饭局、沙龙等多人轮流发言的手机实录按人组织,串成一条每个人都出场的竖版主片,或每位嘉宾各出一条单人短片,逐人落实聊天记录里的不露出与删词要求;一人受访不走这里。Use when every guest must appear in a group dinner or salon film, or each needs a short.

Best for
Video Creation
Current version
v0.5.0

BEFORE YOU START

Check that it can run

Runtime
Portable Agent Skills format. macOS on Apple Silicon. Python 3.9+ in one persistent venv with numpy, opencv-python, pillow, opencc, pypinyin and mlx-whisper; transcription uses two local Whisper models (large-v3 about 3.1 GB, large-v3-turbo about 1.6 GB); FFmpeg/FFprobe; swiftc for the AVFoundation HDR reader and the Vision face and text tools; Node 20+ and Remotion 4 for the starter project. Mixing, card timing and audio QC call lov-media-creator scripts (0.34.0 or later); transcription, BGM and the single-frame scan call lov-interview-media-creator tools that do not read its EDL. Optional OPENROUTER_API_KEY for Lyria BGM (about $0.08 per track, observed 2026-11).
Dependencies
lov-branding-consistency; lov-media-creator(混音、母带、字卡时机与音频质检脚本); lov-interview-media-creator(转写、配乐生成、人声判定、单帧坏帧扫描这几个与 EDL 无关的工具)
Accounts & keys
可选:OpenRouter 的 OPENROUTER_API_KEY(Lyria 配乐,只从环境变量读取,不写入文件)

No Claude or GPT account? You can still try it: a Skill is not tied to one model vendor. Claude is the most dependable, GPT also works, and with only domestic tools you can start from the WorkBuddy option below, though results may differ.

INSTALLATION

Install it in your AI

Yoda, Codex, Claude Code, and other compatible runtimes use the same Skill. Pick your tool for the exact installation steps.

Save your workspace profile once, and Profile-aware Skills can reuse it at runtime.

Configure profile

INSTALL & CONFIGURE

Yoda

Best when you use several agents: install, assign, and invoke skills from one workspace.

  1. 01Open Skills in Yoda, search for “group-interview-media-creator”, then select Install on its detail page.
  2. 02In the target agent settings, set the skill to Auto, Manual only, or Off.
  3. 03Reload the current session and choose it from the composer Skill menu. Yoda inserts the native invocation for the active runtime.
Open Yoda

COMPLETE IN YODA

Yoda syncs the installed skill to configured agents, so you do not have to maintain several directories manually.

FIRST INVOCATION

Use lov-group-interview-media-creator for the following task and validate the result against the skill's acceptance criteria:

Skill declaration: Portable Agent Skills format. macOS on Apple Silicon. Python 3.9+ in one persistent venv with numpy, opencv-python, pillow, opencc, pypinyin and mlx-whisper; transcription uses two local Whisper models (large-v3 about 3.1 GB, large-v3-turbo about 1.6 GB); FFmpeg/FFprobe; swiftc for the AVFoundation HDR reader and the Vision face and text tools; Node 20+ and Remotion 4 for the starter project. Mixing, card timing and audio QC call lov-media-creator scripts (0.34.0 or later); transcription, BGM and the single-frame scan call lov-interview-media-creator tools that do not read its EDL. Optional OPENROUTER_API_KEY for Lyria BGM (about $0.08 per track, observed 2026-11).

GUIDE

Use the method

Join VIP to read the complete guide

The full steps, protected materials, and install package are part of Lovstudio VIP, which unlocks every Skill at once.

See Lovstudio VIP

CASES

Used this skill? Share a real result you are happy with, by writing it yourself or asking an agent to prepare it.

Loading cases…

SKILL CARD / TRUST RECORD

See the evidence before you use it

把一场多人实录(手机竖拍、轮流传麦的饭局或沙龙)按人组织:一条串起全员的竖版主片,每位嘉宾可再各出一条单人短片。每个人提过的不露出、删词要求写成构建时报错的门禁。交付审阅版、字幕、招牌封面、耳听清单、嘉宾确认清单与交付报告,批准后交给发布环节。

CARD v0.5.0

Use case

一场多人聚会用手机录了长录像,想做成一条串起所有人的视频号短片,再给每位嘉宾各出一条愿意自己转发的短片,同时守住每个人提过的露出要求

Owner · Lovstudio.AI

Skill output

可交给发布环节的主片与单人片,带审阅字幕、逐人约束记录和可回读的质检报告。报告写清哪些没验,尤其是没有人耳听审。

隐私门禁:区间、照片框、视频区域,构建时相交即报错; 禁用词命中 0(文本和拼音,整条对白轨和逐段短窗); 时间线引用、字幕、人名条、字卡一致性(check_timeline.py); 字卡时机(card_gate.py)、字幕上屏时间(cue_timing.py)、音效可闻度(sfx_audibility.py); validate_cues、smr_check、母带 −16 LUFS-I / ≤ −3 dBTP; 编码后真峰 ≤ −3 dBTP、编码文件完整解码、bt709 标记、5 fps 联系表、逐帧二维码与单帧坏帧扫描

Requirements

可选:OpenRouter 的 OPENROUTER_API_KEY(Lyria 配乐,只从环境变量读取,不写入文件)

lov-branding-consistency; lov-media-creator(混音、母带、字卡时机与音频质检脚本); lov-interview-media-creator(转写、配乐生成、人声判定、单帧坏帧扫描这几个与 EDL 无关的工具)

Risks & mitigations

要求不露出的人出现在成片里,包括手、袖口和声音 · 禁用区间、照片框、视频区域和禁用词在构建时相交即报错,插入镜头、定格帧、封面帧、片尾帧都查;编码文件再跑逐帧扫描,5 fps 联系表和原图人工看完。门禁只查配置里列出的东西,目视不能省

+8 more

DIMENSIONS

Dimension map

4 dimensions

逐人约束

每个人提过的要求进门禁,成片不违反

首个案例:约束写成禁用区间、画面裁切和禁用词,终版逐帧扫描、联系表人工看完,主片已发表。独立验收针对的是上一版,未发现违反约束;终版新增部分没有重跑,由 agent 自查。门禁不能代替目视。全程没有人耳

片型完成度

一场实录产出主片、单人片、封面与报告

首个案例:主片已发表到视频号,列表回读已声明原创;单人片只到 540p 审阅版,没有终版和发布。作者看过开场和音效后确认出终版,封面否掉一版后选定

交付节奏

能看就发,小改动不重渲全片

首个案例反面教训:开工到发表约 20 小时,一段隔了 7 小时 50 分才发新片。改成先发后每条反馈 11–54 分钟回一版。局部重渲只改不到 100 帧,其余帧逐帧校验不变

打包保真

打包后的脚本和项目原脚本产物一致

用首个案例的素材回放:全部时间线的时间线、字幕、人声清单、对白 WAV、film、seams 逐字节相同,阴性对照都有反应;单人片设计稿到 EDL、配乐方案逐字节相同;主片和单人片的音乐床与混音 WAV 逐字节相同;起步工程静帧逐像素相同。全合成素材从空目录冷启动跑通。没有在真实新项目上完整跑过

VALUE CARD

Value and delivery boundary

Turn one multi-person recording into a vertical film where every guest appears, plus a solo short per guest, with per-guest constraints, captions, music, covers and QC.

Included in Lovstudio VIP

VIP

One price for every VIP Skill

Value anchor

A watchable main film on the day of the recording, without working out the full-cast survey, per-guest privacy gates, frame-level cut builds and music alignment from scratch, or repeating the rework found in the first case.

Usage boundary

Covers the footage survey and cast table, per-guest constraints and privacy gates, frame-level cuts and cut checks, captions and packaging, music and mix gates, sign covers, and delivery QC; third-party AI music fees, font licensing, manual editing, and publishing are separate.

Maintenance

Review when Remotion major versions, the shared contracts with Video Studio and Interview Short, platform delivery rules, or real-use feedback change.

Feedback and comments

Login to comment

DISTRIBUTION

Where you can get it

See the real release status for this Skill on each platform. Links open the exact product page when one exists.

LSPublished

Lovstudio.AI

Get, install, and read the guide on Lovstudio.AI.

You are here

WBNot published yet

WorkBuddy

A public listing is not available yet.

The exact product page will appear here after release.

SPNot published yet

Alipay SkillPay

A public listing is not available yet.

The exact product page will appear here after release.

PAIR WITH

Add these capabilities when needed

DETAILS

Permissions and license

Source visibility
Commercial: packages and docs follow entitlement; public product info stays visible.
Data & permissions
A skill may read local files, access the network, or call third-party models. Review its full steps and commands before running.
License
MIT
videogroup-interviewwechat-channelsremotionfilm-style

NEXT STEP

One membership, every Skill.

Return to the catalog and choose by outcome, evidence, and dependencies. With VIP, every Skill installs directly.

Back to current skill results

Lovstudio.AI / KEEP MAKING

The meaning of life is to create.