agentmods.dev

lecture-recording-captions

lecture-recording-captions is a skill for Claude Code from bam-bam-2/solo-skills. It costs 182 tokens per session (3,006 once invoked), scanned A, original, MIT.

A video-production workflow that turns lecture, live-session, or class recordings into captioned videos with title cards, highlighted speech, selected enlargements, and an ending card. It can prepare the result for YouTube or Discord, but uploading is not the default.

In plain words
What is it for?
Use it to transcribe recordings, add captions, format slides and question-and-answer sections, create a shareable MP4, and prepare distribution notices.
Why use it?
Raw recordings are often hard to follow and may show changing screens, menus, or webcams. This workflow selects the appropriate level of editing and keeps captions when more elaborate visual edits would not fit the source.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the solo-skills plugin — 39 skills shipped together

not rated 367repo +1 6d ago A scan 182 tokens original MIT

Good fit Use it to transcribe recordings, add captions, format slides and question-and-answer sections, create a shareable MP4, and prepare distribution notices.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/bam-bam-2/solo-skills/lecture-recording-captions
View source ↗ bam-bam-2/solo-skills get100.co.kr
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add bam-bam-2/solo-skills --skill lecture-recording-captions
Clone the repo
git clone --depth 1 https://github.com/bam-bam-2/solo-skills

Made for: Claude Code.

Or install solo-skills, the plugin that ships this one along with the rest of its 39 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for lecture-recording-captions

README.md
[![agentmods](https://agentmods.dev/badge/skills/bam-bam-2/solo-skills/lecture-recording-captions/github.svg)](https://agentmods.dev/skills/bam-bam-2/solo-skills/lecture-recording-captions)
Your own site
<a href="https://agentmods.dev/skills/bam-bam-2/solo-skills/lecture-recording-captions"><img src="https://agentmods.dev/badge/skills/bam-bam-2/solo-skills/lecture-recording-captions/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for lecture-recording-captions

Your own site · 80×15
<a href="https://agentmods.dev/skills/bam-bam-2/solo-skills/lecture-recording-captions"><img src="https://agentmods.dev/badge/skills/bam-bam-2/solo-skills/lecture-recording-captions.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 182 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,006 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00182 $0.03006
Opus 5.5 $0.00073 $0.01202
Sonnet 5.5 $0.00036 $0.00601
Haiku 4.5 $0.00018 $0.00301

Measured 1mo ago against content hash 2a92063ac0bd, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

lecture-recording-captions scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.

The scan reads SKILL.md. This mod also ships 9 executable files (scripts/add_words.py, scripts/apply_fix.py, scripts/collect_tr.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/lecture-recording-captions/SKILL.md · 100 lines

How it starts

The opening of the file, as written. The whole thing — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.

이 스킬이 시스템에 하는 일 (설치 전 확인)

  • 로컬에서 ffmpeg과 음성 인식 모델을 돌립니다. CPU·디스크를 많이 씁니다.
  • 녹화 파일을 읽고 자막·영상 파일을 새로 만듭니다. 원본은 수정하지 않습니다.
  • 유튜브 업로드 단계는 별도 인증이 필요하고, 기본값은 업로드하지 않는 것입니다.

강의 녹화본 자막판

결과물: 1920x1080 30fps mp4 하나. 1부(슬라이드 발표) + 2부(질의응답 웹캠) + 엔드카드. 겟백 팔레트(#0d0d0d / 라임 #D4F000 / 퍼플 #7B2FFF), 자막 나눔고딕 ExtraBold, 타이틀·일러스트 배달의민족 주아체(assets/fonts/BMJUA_ttf.ttf, 상업 사용 가능, 엔드카드에 출처 표기).

무거운 단계(전사·교정·렌더)는 전부 맥미니에서 돌린다(macmini-offload 스킬). 맥북에서는 크롭·축소 전처리만 videotoolbox로 한다.

먼저 정할 것 — 판매용인가 기수 공유용인가

두 경로가 갈린다. 소스 화면이 슬라이드로 고정되어 있는지 먼저 본다(0단계 콘택트시트).

판매용 풀 파이프라인 기수 공유용 경량
소스 조건 발표 내내 슬라이드 고정 화면이 슬라이드·브라우저·웹캠으로 계속 바뀜
포함 자막 + 타이틀카드 + 확대 + 일러스트 + PIP 자막 + 엔드카드만
걸리는 시간 반나절 이상 2~3시간

기수 세션 녹화본은 보통 줌 창 전체를 녹화해서 맥 메뉴바·독·디스코드 사이드바가 다 찍힌다. 이런 소스에 슬라이드 구간표를 만들면 타이틀카드가 엉뚱한 화면에 뜨고 확대가 엉뚱한 데를 잡는다. runs를 통짜 하나([[0, 끝, 1]])로 두고 titles·zoom·illustrations를 빈 객체로 두면 자막과 엔드카드만 적용된다. 이 조합은 2026-09-07 7기 OT에서 실측됐다.

손실 없이 빠지는 것과 아닌 것을 구분한다. 자막은 복습 가치가 크므로 반드시 넣고, 타이틀카드·확대는 화면이 고정된 소스에서만 값을 한다.

절차

0. 소스 파악과 싱크

  1. 영상 소스와 음성 소스를 분리해 적는다. 맥 화면기록은 오디오가 없다.
  2. 두 파일을 10분 간격으로 샘플링해 콘택트시트로 본다. 슬라이드가 언제 끝나는지, 화면공유가 언제 꺼지는지, 웹캠 타일이 어디 있는지, 참가자 이름·채팅이 어디 찍히는지 표시한다.
  3. 오프셋은 "슬라이드 제목을 읽는 발화"와 그 슬라이드의 등장 시각으로 두 군데 이상 맞춰 확정한다. 이후 모든 시각은 음성(전사) 타임라인으로 통일한다.
  4. 실제 발화 끝은 볼륨 측정으로 확인한다. 전사가 "하하하"만 찍혀도 말이 이어지고 있을 수 있다.

1. 전처리 (맥북, videotoolbox)

필요 구간만 잘라 뽑는다. 원본은 보내지 않는다.

  • base.mp4 슬라이드 크롭 → 1920x1080 30fps, -c:v h264_videotoolbox -b:v 3500k
  • face1.mp4 발표 중 얼굴 타일 크롭(원본 해상도 유지)
  • bam2.mp4 질의응답 웹캠 타일 크롭 → 1440x810
  • audio.m4a 음성 -vn -c:a copy
  • 마지막 슬라이드 PNG를 1920x1080으로 s18.png처럼 준비(공유 해제 구간 덮기용) ssh localhost로 nohup 실행하고, 중단할 땐 pkill -f '[p]rep.sh' 형식으로.

2. 전사 (맥미니)

줌 클라우드 녹화본이면 전사·교정 단계를 통째로 건너뛸 수 있다. 줌이 audio_transcript.transcript(WEBVTT)를 같이 주고, 한국어 품질이 whisper 못지않다(2026-09-07 실측: 830큐, 밤밤 발화 정확). 절차는 references/zoom-cloud-source.md 참고.

  1. scripts/vtt_to_segs.py <transcript> segs_final.json — WEBVTT → 세그먼트(긴 발화는 문장 단위로 분할)
  2. 고유명사 교정 — 줌 전사도 브랜드명은 틀린다("갯배/겟배" → "겟백", "칠 기" → "7기"). 정규식 치환으로 한 번에 고친다.
  3. scripts/add_words.py segs_final.json — 필수. make_video.py는 어절 하이라이트를 위해 s['words']를 요구하는데 줌 전사는 문장 단위라 없다. 글자 수 비례로 어절 시간을 배분한다. 이걸 건너뛰면 KeyError: 'words'로 죽는다.

Read the full file on GitHub · 100 lines

Files

What ships with it

15 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 1mo ago First seen · 100 lines · 182 tokens per session scan A 2a92063ac0bd

Subscribe to this mod's changes

lecture-recording-captions is a skill published in the GitHub repository bam-bam-2/solo-skills (367 stars, last pushed 6d ago), licensed MIT. It adds 182 tokens to every session and 3,006 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-14.

Claude CodeCreative and mediaLearning and teachingProductivityagent-skillsai-agentautomationkorean
Related

Other skills, from other repositories

summarize-yonsei-today

A read-only daily briefing for Yonsei students, combining accessible information about classes, scholarships, mileage, graduation, teaching progress, LearnUs, attendance, and campus tasks.

mrcha033/yonsei-skills · 73 tokens

founder-coach

Personal leadership development for founders and first-time CEOs. Covers founder archetype identification, delegation frameworks, energy management, CEO calendar audits, leadership style evolution, blind spot identification, imposter syndrome, founder mental health, and succession planning. Use when a founder feels…

alirezarezvani/claude-skills · 101 tokens

cs-onboard

Founder onboarding interview that captures company context across 7 dimensions. Invoke with /cs-onboard setup for initial interview or /cs-onboard update for quarterly refresh. Generates /.claude/company-context.md used by all C-suite advisor skills. Use when setting up the C-suite advisors for the first time, or when…

alirezarezvani/claude-skills · 88 tokens

subscription-setup

A setup workflow for creating recurring medical-learning subscriptions after a doctor’s profile has been registered.

TencentCloud/Octop · 80 tokens

study-plan

Build a weekly and daily study schedule from the goal and the available time, with method notes and adjustment rules, delivered as an Excel workbook.

genspark-ai/genoffice · 30 tokens

scaffold-exercises

A directory scaffold for course exercises, with sections, exercise folders, and optional problem, solution, or explanation versions. It creates the required Markdown files and checks the result with the repository's linter.

vinvcn/mattpocock-skills-zh-CN · 46 tokens