video-to-notes CLI
A CLI and Codex skill turning YouTube and Bilibili videos into transcripts and timestamped summaries, with Notion export.
- 113 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The installer puts dependencies and a Whisper model on your machine.
- Publishing to Notion sends the transcript and summary to that service.
Install
- Works in
Manual install
git clone https://github.com/KIRVO-REPORTING/video-to-notes.git
cd video-to-notes
./install.shCommands from the quick install section for macOS and Linux. Windows PowerShell has `install.ps1`. The Codex skill is copied separately from `codex-skill/`.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The installer puts dependencies and a Whisper model on your machine.
- Publishing to Notion sends the transcript and summary to that service.
This is third-party code. Review the repository files before installing.
What it does
video-to-notes downloads video captions through `yt-dlp` and runs local Whisper only when no suitable captions exist. The transcript becomes an HTML report with clickable timestamps and a summary tied to the source. The `process`, `finalize`, `publish-notion` and `publish-obsidian` commands take the result into a knowledge base, and `serve --open` opens a local dashboard with all reports. The skill in `codex-skill` teaches Codex to call the CLI and write a summary grounded in the transcript.
Who it is for. Students, researchers and writers who take notes from long lectures, interviews and talks on YouTube and Bilibili.
Good fit when
- A two-hour lecture needs notes with timestamps.
- A video with no captions must be transcribed locally.
- Video notes must land in Obsidian or Notion.
- An agent summarizes a clip and must cite minutes.
Not a fit when
- You want a verbatim translation of the whole video.
- The video blocks downloads and no workaround is intended.
- The machine lacks resources for a local Whisper model.
Example request
Make notes from this YouTube video with timestamps and put the report into Obsidian.Limitations
Installation runs the `install.sh` or `install.ps1` script from a clone and prepares Python, `ffmpeg` and a Whisper model, which takes time and disk space. The skill is documented for Codex, so other clients need a manual install path. Some videos need browser cookies via `--cookies-from-browser`, which hands a session to `yt-dlp`. A model writes the summary, so check it against timestamps. YouTube and Bilibili need only a link, while other sites work if `yt-dlp` supports them.
How to disable. Delete `~/.codex/skills/video-to-notes`, the clone directory with its `.venv` and the reports in the workspace. Revoke the Notion token in integration settings.
FAQ
When does Whisper run?
Only when a video has no suitable captions. `--force-transcribe` forces transcription and `--no-transcribe-fallback` forbids it.
How do I send a report to Notion?
Add `--publish-notion` to `finalize` or run `video-to-notes publish-notion` for a finished folder. If Notion is connected as an MCP the agent can use it.
Is the old alias still there?
Yes, `ytlt` works as an alias, but the docs use `video-to-notes`.
Where are reports stored?
In the workspace set by `--workspace`. It holds the transcript, an HTML report and an index for the local dashboard.
What does `doctor` check?
The environment: Python, `ffmpeg`, `yt-dlp` and the selected Whisper backend. Run it before the first processing.
Discussion
No questions or reviews yet.
Related
CLI toolsA skill that fetches YouTube captions and builds an HTML report with takeaways, timestamps and an embedded player
Skill that syncs favorite Douyin videos into a local Markdown or Obsidian base with speech transcripts and a daily digest.
A skill that turns Douyin videos and photo posts into topic notes in Obsidian with local transcription and an index.
Turns a Bilibili or YouTube link or a local file into a transcript, a short abstract and study notes in Markdown
A CLI that keeps agent skills in a Notion database: browse a shared store, install the ones you need, sync and publish for a team
A Codex skill builds an Obsidian vault from your lectures, problem sets and exams with worked derivations and graded practice.
README in short
Captions come first and local Whisper is only a fallback. Reports live in your workspace, and Notion and Obsidian export stay optional.