Lore

Lore is the long-form prose attached to an entity — its history, description, and detail. Where properties capture structured facts, lore is where the writing lives.

Lore vs. properties

A rule of thumb for what goes where:

• Properties hold the structured facts you’ll filter or sort on — dates, statuses, amounts, emails. They’re what an agent’s property filters query.
• Lore holds the prose — history, meeting notes, reasoning, summaries — found by meaning-based search.

When you’re unsure, default to lore: semantic search will still surface it, and a fact you later want to filter on is easy to promote to a property. Keep an entity’s short description to a line or two and put anything longer in lore.

Reading lore in the portal

Open any entity and its lore appears as titled sections in the main column, with a count of how many lore entries it has. Each entry shows when it was last edited and by whom.

Lore powers search

Lore is automatically indexed for meaning-based search, so it can be found by what it says rather than by exact keywords. This is what lets a connected AI tool answer questions like “which people are described as mentors?” even when none of those words appear verbatim. See Search and discovery.

Writing lore

Lore is created and edited through a connected AI tool with read-write access (signed in or via a read-write key). You can have an agent draft it, expand it, or turn rough notes into structured entries — and the result shows up on the entity in the portal right away. For example:

Write a two-paragraph history for the group "Northside Book Club",
covering how it started and what it does now, and attach it as lore.

See Connect an AI tool for more.