The Realm Context is a free-text field where you describe your realm's purpose, theme, and subject matter. This text is included in every AI prompt sent during analysis, giving the language model essential context about your archive.
For example, if your realm documents field recordings from the Scottish Highlands, your realm context might describe the geographic focus, the types of signals you collect, and the themes you care about. The more specific your context, the more relevant and grounded the AI's analysis will be.
The Holder Name field controls how the AI refers to you in its analysis output. This is the name that appears when the AI writes about the person who captured or curated a signal. Set it to whatever name or identity you want the AI to use.
Analysis runs in two passes — surface and structure — each generating different analytical fields. The field selection settings let you choose which fields to include in each pass.
Not every field is relevant to every realm. By selecting only the fields that matter to your archive, you reduce noise in the analysis output and focus the AI's attention on what you care about. Deselected fields are excluded from the AI prompt entirely, saving tokens and improving relevance.