메뉴
Pull a YouTube video's transcript and reshape it into summaries, chapters, threads, or blog posts. Use when: "перескажи это видео", "сделай конспект youtube", "вытащи субтитры из видео", "summarize this YouTube video", "get the transcript", "turn this video into a blog post"
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Finds and downloads free, legally available audiobooks — public-domain recordings from LibriVox and the Internet Archive — via curl + jq, no API key needed. Saves into /Volumes/SanDisk/Audiobooks/ (syncs to phone) when mounted, else the current topic folder, and sends the file(s) via the bot MCP `send_document`. Does not touch piracy sites that host copyrighted commercial audiobooks without a license — that's out of scope. Use when: "скачай аудиокнигу", "найди бесплатную аудиокнигу", "скачай книгу в аудио", "где послушать книгу бесплатно", "хочу аудиокнигу", "download an audiobook", "find a free audiobook", "get me the audio version of [book]"
Add, edit, or remove scheduled tasks AND reminders (scheduled Telegram messages) in the dotfiles cron: picks the right file, validates the schedule, wires logging/locking, reinstalls crontab via init-cron-jobs.sh. New private entries go into the dotfiles_private overlay (~/projects/dotfiles_private/cron/cron_jobs/); shareable jobs into public ~/projects/dotfiles/cron/cron_jobs/. Use when: "добавь крон", "запланируй задачу", "поставь в крон", "запиши в крон", "убери из крона", "schedule a task", "add a cron job", "add to crontab", "remove cron job", "напомни мне", "поставь напоминание", "напоминай каждый", "remind me", "set a reminder", "ping me at".
Pick a typeface for a specific job — book, newspaper, infographic, logo, invitation, screen UI, deck — using Julian Hansen's "So You Need A Typeface" decision tree, then ground the pick in what can actually be licensed and shipped: free substitutes, pairing partner, CSS stack, Cyrillic coverage. Use when: "подбери шрифт", "какой шрифт выбрать", "шрифт для логотипа", "шрифт для книги", "шрифт для сайта", "пара шрифтов", "с чем сочетать шрифт", "чем заменить Helvetica", "бесплатный аналог шрифта", "choose a typeface", "what font should I use", "font pairing", "free alternative to <font>", "pick a font for this landing page". Skip for: copying a specific brand's look (popular-web-designs), slide layout and sizes (powerpoint), installing font files, generating specimens.
Run local GGUF models with llama.cpp (CPU/Apple Silicon/CUDA/ROCm), pick the right quant, and discover GGUF repos on the Hugging Face Hub. Use when: "запусти модель локально через llama.cpp", "подбери квант GGUF", "найди GGUF на huggingface", "run a local GGUF model", "which quant should I use", "serve a model with llama-server"
Build a "design board" style collage out of a batch of photos the user just sent: cuts each object out as a transparent PNG, picks a background colour from the room photo, then spawns 10 parallel agents that each design a distinct collage variant using Python + PIL. Sends back a thumbnail grid so the user picks the variant they like, then ships the full-resolution file. Use when user sends 2+ photos in a row and asks any of: "сделай коллаж", "собери коллаж", "сделай мудборд", "несколько вариантов коллажа", "10 вариантов", "сделай дизайн-борд", "make a collage", "build a moodboard", "design board", "несколько коллажей", "разные раскладки". Single-photo edits go to the `image-edit` skill, not here.
Transcribe a Synthesia-style piano YouTube video (falling note bars on a virtual keyboard) into a clean MIDI, a real-grand-piano mp3, engraved sheet music (PDF + editable MusicXML), and a Guitar Pro (.gp5) file. Fuses audio transcription (Spotify basic-pitch) with computer-vision reading of the falling bars for accurate pitches and frame-accurate timing, adds audio-derived velocity dynamics and a sustain-pedal lane, then engraves with key/tempo detection. Use when: "разбери видео пианино в ноты", "пианино в ноты", "пианино в миди", "сделай ноты из видео", "ноты по видео с ютуба", "сделай midi из этого пианино", "видео в ноты", "сделай guitar pro из видео", "synthesia to midi", "piano video to sheet music", "transcribe this piano youtube video", "falling notes to midi", "youtube piano to pdf"
| name | youtube-content |
| description | Pull a YouTube video's transcript and reshape it into summaries, chapters, threads, or blog posts. Use when: "перескажи это видео", "сделай конспект youtube", "вытащи субтитры из видео", "summarize this YouTube video", "get the transcript", "turn this video into a blog post" |
Use when the user shares a YouTube URL or video link, asks to summarize a video, requests a transcript, or wants to extract and reformat content from any YouTube video. Transforms transcripts into structured content (chapters, summaries, threads, blog posts).
pip install youtube-transcript-api
SKILL_DIR is the directory containing this SKILL.md file. The script accepts any standard YouTube URL format, short links (youtu.be), shorts, embeds, live links, or a raw 11-character video ID.
# JSON output with metadata
python3 SKILL_DIR/scripts/fetch_transcript.py "https://youtube.com/watch?v=VIDEO_ID"
# Plain text (good for piping into further processing)
python3 SKILL_DIR/scripts/fetch_transcript.py "URL" --text-only
# With timestamps
python3 SKILL_DIR/scripts/fetch_transcript.py "URL" --timestamps
# Specific language with fallback chain
python3 SKILL_DIR/scripts/fetch_transcript.py "URL" --language tr,en
After fetching the transcript, format it based on what the user asks for. If the user did not specify a format, default to a summary.
Group by topic shifts, output a timestamped chapter list:
00:00 Introduction — host opens with the problem statement
03:45 Background — prior work and why existing solutions fall short
12:20 Core method — walkthrough of the proposed approach
24:10 Results — benchmark comparisons and key takeaways
31:55 Q&A — audience questions on scalability and next steps
A 5-10 sentence overview covering the video's main points, key arguments, and conclusions. Written in third person, present tense.
Chapters with a short paragraph summary for each:
## 00:00 Introduction (2 min)
The speaker introduces the topic of X and explains why it matters for Y.
## 02:15 Background (3 min)
A review of prior work in the field, covering approaches A, B, and C.
Numbered posts, each under 280 chars:
1/ Just watched an incredible talk on [topic]. Here are the key takeaways: 🧵
2/ First insight: [point]. This matters because [reason].
3/ The surprising part: [unexpected finding]. Most people assume [common belief], but the data shows otherwise.
4/ Practical takeaway: [actionable advice].
5/ Full video: [URL]
Full article with:
Notable quotes with timestamps:
"The most important thing is not the model size, but the data quality." — 05:32
"We found that scaling past 70B parameters gave diminishing returns." — 12:18
--text-only --timestamps.--language to get any available transcript. If still empty, tell the user the video likely has transcripts disabled.--language to fetch any available transcript, then note the actual language to the user.pip install youtube-transcript-api and retry.