SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsanything-to-notebooklm

anything-to-notebooklm

Design & creation

View repository

Updated 10-15 20:05

About this skill

Multi-source content → NotebookLM Processor

Obtain content from multiple sources, upload to NotebookLM to generate various outputs; or browse, query, and ask questions about existing notebooks.

Dependencies:notebooklm-py（CLI v0.3.3+）、markitdown(PPTX/XLSX/EPUB conversion only). First-time use requires notebooklm login。

Decision Tree

用戶請求
├── 想找/瀏覽/查詢/下載既有 notebook 或成品？
│   └── Workflow B：瀏覽 → 閱讀 → 提問
├── 有內容要上傳 / 要生成新東西？
│   └── Workflow A：上傳 → 生成 → 下載
└── 不確定 → 詢問用戶

Workflow A: Upload → Generate → Download

Step 1: Identify content source

Input characteristic Handling method
YouTube URL source add "URL" → source wait <id>
Web page URL source add "URL" → source wait <id>
Google Drive document source add-drive <file_id> "title"
Local .pdf/.docx/.md/.csv/.txt source add "filepath" → source wait <id>
Local .pptx/.xlsx/.epub markitdown "file" -o "$TEMP/converted.md" → source add
Local image/audio source add "filepath" → source wait <id>(automatic OCR/transcription)
Plain keywords source add-research "query"(see below)

Step 2: Create notebook + add sources

notebooklm create "筆記本標題"
notebooklm source add "來源"          # 上傳來源
notebooklm source wait <source_id>    # 等待處理完成
  • A notebook can have at most 50 sources; all source wait before generating
  • Confirm before upload: inform the user of the notebook name about to be created; execute after confirmation

Step 3: Generate

Default behavior is non-blocking(except for mind-map). For interactive scenarios, add --wait to wait for completion.

User intent Command
Podcast/audio generate audio --wait
PPT/slides generate slide-deck --wait
Mind map generate mind-map(synchronous, waits automatically)
Quiz/question generation generate quiz --wait
Report/summary generate report --wait
Video generate video --wait
Infographic generate infographic --wait
Flashcards generate flashcards --wait
Data table generate data-table "description" --wait

User did not specify what to generate → only upload, do not generate; wait for further commands.

For each type's format/style/length parameters → see references/generate-options.md。

Step 4: Download

notebooklm download <type> ./output.<ext>

Common options:--latest(latest output),--format <fmt>(output format),--all(download all).

Full download options → see references/generate-options.md。

AI research search

When the user provides plain keywords, use NotebookLM's built-in research feature:

notebooklm source add-research "query" --mode deep --from web --import-all
notebooklm research wait    # 等待研究完成
  • --mode fast(fast) / deep(deep)
  • --from web(web) / drive（Google Drive）

Workflow B: Browse → Read → Ask

Browse and switch

notebooklm list                              # 列出所有 notebook
notebooklm use <id>                          # 切換（支援部分 ID 匹配）
notebooklm status                            # 當前上下文
notebooklm rename "新名稱"                   # 重新命名

Source management

notebooklm source list                       # 來源清單
notebooklm source fulltext <id>              # 完整索引文本
notebooklm source guide <id>                 # AI 摘要

Ask and converse

notebooklm ask "問題"                        # 對整個 notebook 提問
notebooklm ask "問題" -s <source_id>         # 指定來源提問
notebooklm ask "問題" -c <conversation_id>   # 繼續既有對話
notebooklm ask "問題" --save-as-note --note-title "標題"  # 答案存為筆記

History and outputs

notebooklm history                           # 對話紀錄
notebooklm summary                           # notebook 摘要
notebooklm artifact list                     # 已生成成品清單
notebooklm note list                         # 筆記清單

Share

notebooklm share public --enable             # 公開分享
notebooklm share add user@email.com          # 邀請特定用戶

NEVER List

  • NEVER use source add --wait — this flag does not exist. Use source wait <id> instead
  • NEVER assume generate blocks — it is async by default (except mind-map). Add --wait for interactive use
  • NEVER use --style retro — the correct value is retro-print
  • NEVER use 1-based slide index — revise-slide --slide 0 = first slide (zero-based)
  • NEVER create a notebook without confirming the name with the user first
  • NEVER assume audio downloads are .wav — the actual format is .mp4 container
  • NEVER poll artifact status in the main conversation for extended periods — use a subagent or inform the user of estimated wait time

Integration and handoff

  • yt-search: YouTube URL directly source add; if you already have a subtitle text file, you can also directly source add
  • markitdown: only for PPTX/XLSX/EPUB (formats not directly supported by NotebookLM). PDF/DOCX/MD/CSV/TXT/image/audio directly source add
  • Obsidian vault: after reading the note content source add file path
  • crawl4ai: for anti-scraping websites (X.com, etc.), first use crawl4ai to scrape, then source add local file

Language settings

notebooklm language set zh_Hant    # 繁體中文
notebooklm language set ja         # 日文

You can also add to individual commands --language zh_Hant。

Temporary file cleanup

After generation is complete and the user confirms receipt of the output, clean up temporary files:rm -f "$TEMP"/converted.md

References

What you need Where to look
Generation/download parameter details references/generate-options.md
Complete workflow examples references/examples.md
Windows setup references/windows-setup.md
Troubleshooting references/troubleshooting.md