Insight and output

Assistant

A private chat about your own figures, running entirely on your device. Currently in beta.

The Assistant answers questions about the data Monestra already holds — "what am I spending on subscriptions?", "how has my net worth changed?" — by narrating figures the app has already computed.

What it can see#

Only visible modules. Hide a module and the Assistant stops being able to discuss it.

Within those, it reads two levels of detail: the same rollup the Dashboard gets, and a deeper item-level view — line items, category breakdowns, pace. It also reads value history, which is what lets it answer questions about change over time.

What it will not do#

• It will not invent a number. It is instructed to narrate figures the app computed and nothing else.
• It will not advise you. Same rule as everywhere in Monestra — it describes, it does not recommend. See What Monestra is.
• It will not show amounts while masking is on. Figures come back as ••••, exactly as they do everywhere else.
• It will not fake a reply. If the on-device model is unavailable, you get a specific explanation of why, not a canned answer.

When the model is unavailable#

The Assistant needs a device that supports Apple Intelligence, with it enabled and the model downloaded. Where that is not true, the Assistant says which of those is the problem — unsupported device, Apple Intelligence off, model still downloading, or an OS that is too old — rather than failing vaguely.

Beta#

The Assistant ships behind a beta badge, in the sidebar, in Settings ▸ Modules and on its own welcome screen. It is hideable like any other module.