Build your own entity graph layered over a study room's canonical world. Add your own characters, places, and connections. Let your Personal Assistant help you find gaps and contradictions in your theory.
A personal entity graph that inherits the room's world
How to enable and configure MyWorld access
How to create and build your personal world
Characters, places, timelines, and how they connect
Ask your Personal Assistant about your world
Find missing entities and contradictions
Walkthrough: building a suspect theory in a murder case
Every study room has a canonical world— a shared set of characters, places, timeline events, objects, and context entries controlled by the room's owner and moderators. This is the “official” knowledge graph that all members can see and Sage (the room's AI) uses to answer questions.
MyWorld is your personal version of that world. It starts as a read-only copy of the canonical entities, but you can add your own entities, draw your own connections, and build your own theory — entirely private to you.
As a room owner or moderator, you control who can create a MyWorld in your room. The feature is disabled by default — you must explicitly enable it.
/study-rooms/[your-room]/settings/world.MyWorld is a premium feature. In free Starter meeting rooms, the World and MyWorld features are locked behind a Reader+ upgrade prompt. Members will see the MyWorld tab with a lock icon and an upgrade CTA.
Once enabled, members will see a 🧭 My Worldtab in the room's Bookshelf dropdown navigation. They can also find a button on the World page itself. Encourage adoption by:
If the room owner has enabled MyWorld, you can create your own personal world in just a few clicks.
There are two ways to add entities to your MyWorld:
Hover over any entity in your My World column and click Edit to open the entity editor. You can change the name, type, description, and confidence level. Click Delete to remove the entity and all its relationships.
MyWorld uses the same five entity types as the canonical world:
People, suspects, witnesses, family members. The central actors in your world.
Locations, crime scenes, destinations, addresses. Where things happen.
Events, dates, sequences. What happened and when.
Evidence, weapons, heirlooms, documents. Physical items of significance.
Rules, hypotheses, constraints. The framework for your investigation.
Each entity can have a confidence level that signals how certain you are about it:
Relationships connect entities in your MyWorld. They are typed links with a label — for example:
To create a relationship, click + Add Relationship in the Relationships section. Pick a source entity, type a label, and pick a target entity. Relationships help the AI understand the structure of your theory.
Your Personal Assistant (PA) is your private AI that follows you everywhere on the platform. When you open the PA from inside a study room where you have a MyWorld, it becomes aware of both the canonical world and your personal world.
/homeroom/pa/settings. Call it “Watson,” “Holmes,” or whatever fits your investigation style.| Feature | Sage (Room AI) | Your PA |
|---|---|---|
| Who sees it? | Everyone in the room | Only you |
| What does it know? | Canonical world only | Canonical world + your MyWorld |
| Where does it live? | Room chat panel | Popup (follows you everywhere) |
| Can I rename it? | No — set by room owner | Yes — at /homeroom/pa/settings |
| Cross-room awareness? | Connected rooms only | Yes — sees all your MyWorlds across all rooms |
MyWorld includes AI-powered analysis that helps you find gaps, contradictions, and missing connections in your personal world. Each analysis costs creditsthrough the platform's usage-based billing system.
The first available analysis type is ENTITY_GAP. It compares your MyWorld to the canonical world and suggests entities you might be missing:
Additional analysis types are planned for future releases:
Let's walk through a realistic scenario: you're a member of a mystery room investigating the murder of restaurant owner Carlo Bianchi. The room's canonical world has a full set of entities — suspects, witnesses, timeline events, evidence. But you have a hunch that the room's primary theory is wrong. You think Serena Voss, the business partner, is being overlooked.
Navigate to your study room and click the My World tab. Click Create My World and give it a title: “The Partner Did It”.
Your MyWorld loads with the canonical world's entities visible as a reference on the left — but your personal world on the right is empty. This is your blank canvas.
Relationships are where your theory takes shape. Each connection is a statement about how entities relate.
The canonical timeline places Carlo's death between 10pm and midnight on January 14. Your theory requires a different sequence.
You have a solid suspect theory — 8 entities and 7 relationships. Now ask the AI if it spots gaps.
Open your PA popup from inside the study room. It now knows about both the canonical world and your MyWorld.
💬 “What contradictions exist between my world and the main world?”
The main world places Serena at her apartment during the time of death. Your world places her in the back office at 11:28pm — directly contradicting the alibi she gave to police. The canonical timeline has a 2-hour gap (10pm–midnight) for Serena; your world fills this gap with security footage and staff testimony.
💬 “Does my timeline make sense?”
Your timeline is internally consistent. The critical window is 11:28pm (Serena enters) to 11:52pm (estimated exit). Carlo was last seen at 10:45pm and found at 12:15am. Your theory puts Serena on scene during the 23-minute window when Carlo was likely killed. One gap: you don't have a clear account of what happened between 11:52pm and 12:15am. Consider finding evidence of Serena's movements during that period.
💬 “Who am I missing as a potential accomplice?”
Based on your relationship graph, the dishwasher who last saw Carlo at 10:45pm is not in your MyWorld. They could confirm or contradict the timeline. Also, the restaurant's security guard is not in the canonical world — you may want to create this entity yourself if you suspect a third party was involved.
After an hour of focused work, your MyWorld now contains:
Your theory — that Serena Voss, motivated by financial gain, was present at the scene during the critical window and had means and opportunity — is now a structured, analyzable world graph. The PA has validated its internal consistency and flagged gaps for further investigation.
And here's the key: none of this is visible to other room members. The canonical world is unchanged. You can choose to share your theory when you're ready — or keep refining it until you have a compelling case to present.
The room owner has not enabled MyWorld, or your role does not meet the access threshold. Ask the room owner to check the MyWorld Access setting at the room's World settings page.
Make sure you're opening the PA from inside the study room— the PA detects room context based on which page you're on. If you open the PA from the dashboard or another page, it won't have room context.
Also, use “my” language in your questions — “characters in myworld” not just “characters in the world.”
AI analysis costs credits. You can top up your credit balance through the billing dashboard. Your MyWorld and all its entities are preserved — you just need credits to run new analyses.
Your MyWorld is preserved when you leave a room and restored when you rejoin. If you were removed from the room, your MyWorld is soft-deleted — it will be restored if you rejoin, but you won't be able to access it while you're not a member.