Interface DomainEntry
interface DomainEntry {
slug: string;
title: string;
summary: string;
whenToSearch: string;
whenToSave: string;
whatNotToSave: string;
tagsDescription: string;
filterTagTypes: readonly string[];
versionTagTypes: readonly string[];
maxTagsPerQuery: number;
}
slug: string;
title: string;
summary: string;
whenToSearch: string;
whenToSave: string;
whatNotToSave: string;
tagsDescription: string;
filterTagTypes: readonly string[];
versionTagTypes: readonly string[];
maxTagsPerQuery: number;
}
Properties
Readonlyslug
slug: string
The handle to pass as domain.
Readonlytitle
title: string
The domain's human-readable name.
Readonlysummary
summary: string
What the domain holds.
ReadonlywhenToSearch
whenToSearch: string
When to search this domain.
ReadonlywhenToSave
whenToSave: string
When to save into it.
ReadonlywhatNotToSave
whatNotToSave: string
What does not belong in it.
ReadonlytagsDescription
tagsDescription: string
The tag vocabulary, and the format it takes.
ReadonlyfilterTagTypes
filterTagTypes: readonly string[]
Tag types that narrow results rather than boosting them.
ReadonlyversionTagTypes
versionTagTypes: readonly string[]
Tag types that carry a version.
ReadonlymaxTagsPerQuery
maxTagsPerQuery: number
How many tags one call may carry. Extra tags are trimmed, not refused.
One memory domain the caller may name.
Most of it is prose the service authored, and briefing renders it as a system prompt so a model is told what you are reading here.
Two fields change what a call does rather than describing the domain. DomainEntry.filterTagTypes is what decides whether a wrong tag costs ranking or costs every result, and DomainEntry.maxTagsPerQuery is a cap that trims — per-domain, which is why it sits here and not on Limits.
See
Tag — the vocabulary these fields describe.