Tana UI changes instructions

Good catch on #video. From the screenshot I can see it extends #source and has fields: Host, Publication, Media Source, Year, Quotes, Highlights, Zettelkasten Entry. It’s mostly documentaries and talks — different from #film (which has Director, Writer, Cast, Producer, Genre, Language). Makes sense to keep both.

Let me also note what I see about your current #author (★) and #film (★):

#author (★) — extends #person (★), has fields: Works, Quotes and Highlights, Related Notes, Website(s). This is well-structured already.

#film (★) — extends #source, has fields: Director, Writer, Genre, Language, Year, Media Source, and Topic(s) ★. Similar pattern to #book before rationalisation.

Phase 1 & 2 — Complete Tana UI instructions

Work through these in order. Each step says exactly what to do in the Tana UI. Estimate: 45-60 minutes total.

Part A — Sever inheritance chains (5 minutes)

A1. Open #book supertag config → Building blocks → remove #source from “Extend other supertags”. (The fields are already on #book independently from the migration you did.)

A2. Open #film (★) supertag config → Building blocks → remove #source from “Extend other supertags”. The fields (Director, Writer, etc.) stay — they’re defined on the tag itself.

A3. Open #podcast episode (★) supertag config → Building blocks → remove #source from “Extend other supertags”.

A4. Open #video (★) supertag config → Building blocks → remove #source from “Extend other supertags”.

A5. Open #note supertag config → Building blocks → remove #input from “Extend other supertags”. (Keep the tag — it’s used in Projects/Areas. Just sever the OCP inheritance.)

Part B — Clean up tag names (10 minutes)

For each of these, open the supertag config and edit the name:

B1. #author (★) → rename to #author B2. #film (★) → rename to #film B3. #podcast episode (★) → rename to #podcast episode B4. #podcast (★) → rename to #podcast B5. #video (★) → rename to #video B6. #person (★) → rename to #person B7. #evergreen (★) → rename to #evergreen B8. #garden card (★) → rename to #garden card B9. #host (★) → rename to #host B10. #literature note (★) → rename to #literature note

For each renamed tag, add a description if missing. Use the descriptions from the ontology reference (e.g. #author: “People who create works — books, articles, films, podcasts, music”).

Part C — Create new tags (15 minutes)

C1. Create #idea

• No extends (independent root)
• Description: “Your own thinking — sparks, plans, reflections”
• Base type: None
• Fields to add:
  • Topics — instance of #topic, multi-value
  • Source — instance of #article, #book, #film, #podcast episode, #video, #tweet (what prompted this idea)
  • Related — multi-value, any node type (other ideas, concepts, claims, sources)
  • Status — options: seed · hypothesis · draft · active · evergreen
  • Media Source — url (optional, hide when empty)

C2. Create #idea/fleeting

• Extends: #idea
• Description: “Sparks, observations, quick captures”
• No additional fields (inherits from #idea)

C3. Create #idea/writing

• Extends: #idea
• Description: “Notes about things you intend to write”
• No additional fields

C4. Create #idea/journal

• Extends: #idea
• Description: “Reflective, time-bound entries, diary”
• No additional fields

C5. Create #claim

• No extends (independent)
• Description: “Assertable, evidential. Can be backed or refuted by evidence.”
• Fields to add:
  • Topics — instance of #topic, multi-value
  • Source — instance of source types (evidence or provenance)
  • Related — multi-value (other claims, ideas, concepts)
  • Status — options: seed · hypothesis · draft · active · evergreen
  • Media Source — url (optional, hide when empty)

C6. Create #tweet

• No extends (independent)
• Description: “Tweets and X posts — short-form social media content”
• Fields to add:
  • Author — instance of #person / #author
  • Topics — instance of #topic, multi-value
  • Media Source — url (link to the tweet)
  • Summary — plain text (your note on why you saved it, hide when empty)

C7. Create #reference note (if it doesn’t already exist as a clean tag)

• No extends (independent)
• Description: “What the author said, distilled in your words”
• Fields:
  • Source — instance of source types
  • Topics — instance of #topic, multi-value
  • Status — options: seed · draft · active · evergreen

Part D — Update existing tags with missing fields (10 minutes)

D1. #film — check and add these fields if missing:

• Director — instance of #person / #author
• Writer — instance of #person / #author (hide when empty)
• Cast — instance of #person, multi-value (hide when empty)
• Producer — instance of #person (hide when empty)
• Confirm existing: Topics, Year, Genre, Language, Media Source, Status

D2. #video — check current fields against your needs:

• Host is already there — good → instance of #person
• Publication is already there — good
• Add Director — instance of #person / #author (hide when empty) if useful for documentaries
• Confirm: Media Source, Year, Topics are present

D3. #podcast episode — check and add:

• Host — instance of #person / #author
• Guest(s) — instance of #person / #author, multi-value
• Podcast — instance of #podcast (the show)
• Confirm: Topics, Media Source, Status are present

D4. #author — review current fields (Works, Quotes and Highlights, Related Notes, Website(s)). These look good for the knowledge branch. No changes needed unless you want to add Topics for subject-area tagging of authors.

D5. #person — check it has at minimum: name (node title), Topics (optional), Media Source (website/profile, optional). This is the lean base — CRM fields live on #contact and its children, not here.

D6. #publisher — if this tag doesn’t exist yet, create it:

• Extends: #organisation
• Description: “Book publishers”
• No additional fields needed beyond what #organisation provides

Part E — Audit node counts via Tana AI Chat (10 minutes)

Open Tana AI Chat and ask these questions. Copy the answers into a note for your Claude Code session:

• “How many nodes have the #thoughts ★ tag?” (or whatever it’s called after renaming)
• “How many nodes have the #fleeting note tag?”
• “How many nodes have the #writing inbox tag?”
• “How many nodes have the #day notes tag?”
• “How many nodes have the #ideas tag?”
• “How many nodes have the #to-sort tag?”
• “Show me 3 example nodes for each”

#idea notes 375, of which 9 simple, plus #idea/writing 120 #idea/fleeting 236 #idea/journal 10

#to-sort is used 32 times, however I often group several nodes and I apply the tag to the parent node. At times I have applied it directly to the node. These may be pure links or text nodes with or without links.

This gives you the migration scope for Phase 3 (bulk retag).

Part F — Set up the #to-sort Text Processing Agent (15 minutes)

F1. Create a new command node in your workspace F2. Set command type: Text processing agent F3. Set node filter: Nodes with tag #to-sort F4. Set tag choices: #article/academic, #article/press, #article/web, #book, #podcast episode, #publication, #tweet, #video, #idea/fleeting F5. Paste the prompt I wrote in my previous message F6. Map the fields to set: Media Source, Author(s), Published in, Topics F7. Test on 2-3 #to-sort nodes you know well — verify classification, title, summary, and field population F8. Add as a button on the #to-sort supertag config (under AI and Commands)

After completing all of this, update your HANDOFF.md with what you did, then you’re ready for Phase 3 (bulk retag in Tana UI) and Phase 4 (Obsidian work with Claude Code).