Automation Rules

Automation Rules are simple “If this happens → then do that” workflows that keep your NotebookLM workspace organized and up-to-date.

The Basic Idea

Each rule has two parts:

  • Trigger — the thing you do (add a tag, move a notebook into a collection, finish an episode, etc.)
  • Action — what Kortex does for you (move a notebook, create notes, generate audio, add to a podcast feed, etc.)

Why Use Automation Rules

  • Less busywork — stop doing the same organizing steps over and over
  • Consistent structure — every project starts the same way
  • Faster workflows — tagging once can trigger multiple steps

Common Examples

  • Tag → Collection: Tag a notebook #work and it automatically gets filed into Office Projects
  • Tag → More Tags: Tag #final and it automatically adds #reviewed + #archive
  • Collection → Audio: Move a notebook into Daily Read and it starts generating an Audio Overview
  • Tag → Podcast: Tag #listenlater and its audio gets added to your podcast feed

Automation Types

Smart Pipelines (Multi-Rule Recipes)

Most people get the best results by combining multiple rules into a single pipeline. Think of it like building a conveyor belt:

Pipeline 1: “Inbox → Processed → Archived” (the universal workflow)

Goal: Every new notebook starts in an inbox, gets processed, and ends up archived with consistent tags.

  1. Create Collections: Inbox, Active, Archive
  2. Smart Sort: Tag #active → move to Active
  3. Smart Sort: Tag #archive → move to Archive
  4. Escalator: Tag #archive → auto-add #reviewed + #done
  5. Template Engine: When moved to Active → create notes: Goals, Key Sources, Next Actions

Pipeline 2: “Tag once → full project setup” (tag macro + structure)

Goal: Apply one tag and the notebook is immediately standardized for your workflow.

  1. Escalator: Tag #client → auto-add #work + #priority
  2. Smart Sort: Tag #work → move to Work Projects
  3. Template Engine: On tag #client (or on entering Work Projects) → create notes: Brief, Deliverables, Meeting Notes, Open Questions

Goal: Save links from many sites but have them land in the right notebook every time.

  1. Domain Router: If URL contains arxiv.org → notebook Papers
  2. Domain Router: If URL contains youtube.com → notebook Videos to Watch
  3. Escalator: When you tag #paper → auto-add #to-summarize + #cite
  4. Auto-Researcher: On tag #to-summarize → generate a Briefing Doc (and/or Flashcards if it’s for studying)

Pipeline 4: “Daily Reading → Auto Audio Playlist” (collection → audio → podcast)

Goal: Turn a folder into a listening queue with minimal effort.

  1. Create Collection: Daily Read
  2. Generator: When notebook enters Daily Read → generate Audio Overview
  3. Pipeline: Collection Daily Read → Podcast Morning Commute
  4. Cleaner: When an episode finishes → remove it from the feed

Best practice: Use the collection as your “single source of truth” queue. Drag notebooks in/out to control what gets produced and queued.

Pipeline 5: “Listen Later” without moving notebooks (tag → podcast)

Goal: Keep notebooks where they belong, but still queue audio for listening.

  1. Create Tag: #listenlater
  2. Podcast Curator: Tag #listenlater → Podcast My Playlist
  3. Cleaner: Finished episode → remove it

Pipeline 6: “Course Notes → Study Materials + Audio” (class automation)

Goal: Every time you file a notebook into a class folder, you get study outputs automatically.

  1. Create Collection: CS224N (or any class)
  2. Template Engine: Entering CS224N → create notes: Lecture Summary, Key Terms, Exam Questions
  3. Auto-Researcher: Entering CS224N → generate Study Guide + Flashcards
  4. Generator: Entering CS224N → generate Audio Overview
  5. Pipeline: CS224N → Podcast CS224N Audio

Pipeline 7: “Meeting Notes Standardizer” (tag → notes template)

Goal: Any notebook tagged as a meeting automatically gets a consistent structure.

  1. Escalator: Tag #meeting → auto-add #work + #action-items
  2. Template Engine: On tag #meeting → create notes: Agenda, Decisions, Action Items, Follow-ups
  3. Smart Sort: Tag #work → move to Work Projects (optional)

Pipeline 8: “Deep Work / Focus mode” (tags become modes)

Goal: Switch a notebook into a mode that triggers multiple automations.

  1. Escalator: Tag #focus → auto-add #noisy-sources-off + #summarize (use whatever tags match your process)
  2. Auto-Researcher: On tag #summarize → generate a Briefing Doc
  3. Podcast Curator: Tag #focus → Podcast Focus Queue (optional)

Pipeline 9: “Client Deliverables Assembly Line” (project folder → outputs)

Goal: When a notebook enters a delivery folder, the outputs you need are generated automatically.

  1. Create Collection: Deliver This Week
  2. Template Engine: Entering Deliver This Week → create notes: Deliverable Outline, Risks, Client Questions
  3. Auto-Researcher: Entering Deliver This Week → generate a Briefing Doc (or other artifact you ship)
  4. Generator: Entering Deliver This Week → generate Audio Overview (for review while walking/commuting)

Pipeline 10: “Topic Router + Filing” (URL routing then tag-based filing)

Goal: Use URL patterns to get things into the right notebook, then tags to file notebooks into the right collections.

  1. Domain Router: If URL contains topic=machine-learning → notebook ML Research
  2. Escalator: Tag #ml → auto-add #paper + #to-summarize
  3. Smart Sort: Tag #ml → move notebook to collection Machine Learning
  4. Auto-Researcher: Tag #to-summarize → generate Briefing Doc

Pipeline 11: “Multi-Playlist System” (one notebook, multiple destinations)

Goal: Use tags to send the same notebook’s audio to different podcast feeds depending on what you need.

  1. Podcast Curator: Tag #commute → Podcast Morning Commute
  2. Podcast Curator: Tag #workout → Podcast Workout Queue
  3. Podcast Curator: Tag #sleep → Podcast Wind Down
  4. Escalator (optional): Tag #commute → auto-add #listenlater

Tip: Keep tags mutually exclusive if you only want one destination at a time, or allow multiple tags if you want cross-posting.

Pipeline 12: “One Tag to Rule Them All” (master trigger tag)

Goal: Apply a single master tag and it triggers a full chain of organization + output creation.

  1. Escalator: Tag #ship → auto-add #work + #priority + #listenlater
  2. Smart Sort: Tag #work → collection Work Projects
  3. Template Engine: On tag #ship → create notes: Scope, Checklist, QA
  4. Generator: On entering Work Projects (or on tag #ship) → generate Audio Overview
  5. Podcast Curator: Tag #listenlater → Podcast Work Review Queue

Pipeline 13: “Two-stage review” (draft → final → archive)

Goal: Make it hard to accidentally archive something before it’s reviewed.

  1. Create Tags: #draft, #ready, #final
  2. Escalator: Tag #final → auto-add #archive + #reviewed
  3. Smart Sort: Tag #draft → collection In Progress
  4. Smart Sort: Tag #ready → collection Review
  5. Smart Sort: Tag #archive → collection Archive

Pipeline 14: “Podcast-first research loop” (listen → tag → artifacts)

Goal: Use listening as the first pass, then automatically generate artifacts for anything worth keeping.

  1. Podcast Curator: Tag #listenlater → Podcast Inbox Audio
  2. Escalator: Tag #keeper → auto-add #to-summarize + #cite
  3. Auto-Researcher: Tag #to-summarize → generate Briefing Doc (or Study Guide)

Pipeline 15: “Personal knowledge base lanes” (topics as collections)

Goal: A clean long-term system where adding a topic tag files it and boots up the right templates.

  1. Create Collections: Health, Finance, Career, Learning
  2. Smart Sort: #health → Health, #finance → Finance, etc.
  3. Template Engine: Entering Health → create notes: Summary, Action Plan, Questions
  4. Auto-Researcher (optional): Entering Learning → generate Flashcards

Good Rule Design (Quick Tips)

  • Start with 1 pipeline — build one “lane” (like Daily Read) and run it for a week
  • Name tags like verbs — #listenlater, #to-summarize, #archive
  • Prefer one main trigger — choose either a tag or a collection move as the main “start” signal
  • Keep rules predictable — if a rule surprises you once, make the trigger more specific