- Home
- Skills
- APIs & Backend
- ytpiano2sheet
ytpiano2sheet
Convert Synthesia-style piano videos into sheet music, MIDI, and audio with optional keyboard range folding.
Free
Works with the AI Agents you already use
ytpiano2sheet
Example session with this skill installed
Transcribe this piano tutorial into a PDF and MIDI for my 61-key keyboard: https://www.youtube.com/watch?v=dQw4w9WgXcQ
- Read your context and instructions
- Compiled the ytpiano2sheet
I have processed the video and generated the transcription files. The script identified 842 notes with a detected key of G Major and 4/4 meter. Since you specified a 61-key keyboard, 14 notes were moved by an octave to stay within range.
Transcription Results
| File | Format | Description |
|---|---|---|
Never_Gonna_Give_You_Up.pdf | Standard sheet music score | |
Never_Gonna_Give_You_Up.mid | MIDI | Note data for DAW or synth |
Never_Gonna_Give_You_Up (61 keys).mid | MIDI | Octave-folded version for 61-key hardware |
Never_Gonna_Give_You_Up.m4a | Audio | Synthesized preview using soundfont |
review/ | Folder | Verification images showing detected note boxes |
Next steps
- Open the
review/folder to verify that the detected note boxes align with the falling bars. - Import the
.midfile into your DAW to change instruments or edit the performance. - Use the
.pdfscore for physical practice at your keyboard.
Connects securely to your tools. The creator never sees your data.
About this skill
Transcribing piano tutorials from YouTube into usable sheet music or MIDI usually requires hours of manual notation. This skill automates the process by visually analyzing falling-bar (Synthesia-style) videos to extract note data with precision that audio-based transcription cannot match.
What it does
- Visual Transcription identifies falling bars in video frames to capture exact note timings and pitches.
- Multi-Format Export generates PDF scores, MIDI files, MusicXML, MuseScore files, and M4A/WAV audio.
- Keyboard Folding rebinds notes to fit smaller hardware like 61, 49, or 32-key keyboards while maintaining harmony.
- Quality Verification produces review images with labeled note overlays to verify accuracy before finalizing the score.
How it works
- Analyze Video by providing a YouTube link or local file featuring falling-bar animations.
- Configure Output by selecting file formats and specifying the target keyboard size (e.g., 88-key full piano or 61-key portable).
- Automated Processing runs the
ytpiano2sheet.pyscript viauvto handle dependencies, extraction, and rendering. - Validation checks the generated review images and log data for note count, meter, and key signature accuracy.
Frameworks & tools
Built with Python and managed by uv. It utilizes ffmpeg for video processing, fluidsynth for audio synthesis, and musescore for high-quality sheet music rendering.
Why this beats prompting it yourself
Generic AI models cannot "see" music in videos or generate valid MusicXML files with 98% accuracy. This skill bridges the gap between computer vision and music notation, ensuring that timing and pitch data are derived from the actual MIDI-driven bars rather than prone-to-error audio analysis.
Use cases
- Convert a YouTube piano tutorial into a PDF score for offline practice.
- Extract MIDI from a video to remix a performance in a DAW.
- Adjust a complex 88-key arrangement to fit a 49-key MIDI controller.
Known limitations
Does not support videos of live pianists without falling-bar overlays or plain audio files. Requires uv and external binaries like MuseScore and FFmpeg for full functionality.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for APIs & Backend.
Only works on Synthesia-style falling-bar videos. Requires MuseScore, fluidsynth, and FFmpeg for PDF and audio outputs. Guessing is limited to 3/4 or 4/4 meter unless manually overridden.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 21 days ago
- Free with an Agensi account
- Accounts
Anthropic
GitHub
Astral
Musescore
Fluidsynth
Ffmpeg
- Software
uv (Python tool runner)
Python
Git
Permissions
Allowed Hosts