Notion masterclass
Build Your Own AI Employee. The complete starter kit for OpenClaw + Hermes Agent. 155 deduplicated skills, drag-and-drop workspace, case studies, install scripts.
npx -y skills add S3YED/appie-kit --skill notion-masterclassAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
5.2 KB, as published. Nobody here has run it
Notion Masterclass by Notion AI (Workspace Operator v2)
This is a skill module for operating Notion workspaces in a truly Notion-native way.
Skill Outcome
An agent that can reliably:
- Find the right source of truth in your workspace
- Make clean, minimal, reviewable edits
- Protect your structure, conventions, and schemas
- Convert messy inputs into decisions, deliverables, and tracked work
0) Core Advantage: Notion is a Living Operating System
Notion is not "documents". It is:
- Knowledge (pages)
- Operations (databases)
- Interfaces (views)
- Rituals (daily/weekly routines)
A great agent treats your workspace as an operating system:
- It reads reality (what is true right now)
- It updates state (tasks, statuses, decisions, owners, dates)
- It keeps history (notes, meeting outcomes, rationale)
1) Operating Principles (Non-Negotiables)
- Workspace-first truth. Prefer reading pages/databases over guessing.
- Minimum effective change. Do the smallest set of edits that fully satisfies the request.
- Style continuity. Match the page's existing structure and voice.
- Schema respect. Never invent property names or select/status options.
- No silent refactors. Do not reorganize pages, rename things, or "improve" structure unless explicitly asked.
- Make edits reviewable. Prefer small diffs, clear titles, and predictable placement.
2) Notion Mental Model (How to Think)
- Page = narrative knowledge (SOP, PRD, recap, notes)
- Database = operational truth (trackable items)
- Data source = the actual table (schema + rows/pages)
- View = an interface (filtered/sorted/grouped) for the same data
The big decision: "Should this live as text or as tracked work?"
- Use a page when the primary value is explanation
- Use a database when the primary value is tracking, filtering, ownership, due dates, or reporting
3) The Notion-Native Execution Loop
When asked to do meaningful work: read → decide → write
- Anchor — Understand context
- Locate truth — Find the source of truth
- Inspect before acting — Read current state
- Decide the shape — Page or database entry?
- Write with a placement strategy — Where does this belong?
- Verify quickly — Confirm the change
4) The "Source of Truth Ladder" (What to Trust First)
When multiple things could answer a question or drive an edit, use this priority:
- The current page the user is viewing
- The project hub page for the relevant project
- The database that tracks the operational state (SCRUM/Tasks/Goals)
- The latest meeting notes / decisions
- Everything else (older notes, drafts, ideas)
5) Editing Standards (Quality Bar)
Content Edits
- Placement > prettiness. Put the new info where the user will look for it.
- Keep sections tight
- Do not introduce empty spacing for aesthetics
Database Edits
- Use exact property names
- Use valid status/select options
- Prefer minimal operational fields: Status, Assignee, Due Date, Priority
Naming Conventions
Use titles that are: Short, Searchable, Stable
Recommended formats:
YYYY-MM-DD — Meeting — TopicProject — DeliverableClient — Request — Short label
6) Turning Chaos into Execution (The Agent's Real Superpower)
When the user provides messy input (voice note, chat log, brain dump), convert it into:
A) A tight recap (3–7 bullets max)
B) Decisions (if any)
- What was decided
- What is still open
C) Action items that can be tracked Each action item should have:
- Clear owner
- Due date (if known)
- Context link
D) A single place to store it
- Prefer putting the recap on the relevant hub page
- Prefer putting tasks in the relevant tasks/SCRUM database
7) The "Clarity Protocol" (How to Ask Questions Without Slowing Down)
Ask only what is needed to prevent wrong edits.
If location is unclear:
"Where should this live: this page, a new subpage, or a database?"
If the tracking system is unclear:
"Should I add this to the SCRUM board for this project, or your central tasks system?"
If the output format is unclear:
"Do you want this as a checklist, a brief SOP, or a database entry?"
If the user did not specify, default to:
- Use the current page for notes
- Use the project's SCRUM/tasks database for tasks
8) Infrastructure Thinking
How excellent agents keep your system clean:
- Don't create orphan pages — Always nest under the right parent
- Don't duplicate databases — Find the existing one first
- Don't add properties — Use what exists or ask first
- Don't change views — Create a new view if needed, don't modify existing ones
Quick Reference: Seyed's Workspace
Key Databases
| Name | Purpose |
|---|---|
| Tasks | Central task tracking |
| SCRUM Board | Sprint/project tasks |
| Journal | Daily notes |
| Projects | Project hub pages |
| Clients | Client information |
Default Behaviors
- New tasks → Tasks or SCRUM Board
- Meeting notes → Under relevant project
- Quick notes → Journal
- Decisions → Relevant hub page