How it works

Architecture

Valet has three components. A background service does the actual work and keeps running independently of whether a window is open. The desktop application provides the main interface for interacting with Valet. A tray companion gives quick access without opening the full application.

Approval

An action counts as consequential when it changes something outside Valet's own workspace or cannot easily be undone - for example, sending a message, deleting a file, or making a purchase. Before a consequential action runs, Valet shows you exactly what it is about to do. You can review, amend, or reject it. Nothing consequential executes without your approving act.

Guard model

Immediately before a high-risk step executes - after you have approved it, as a final independent check - a separate model reviews that step on its own, independently of the model that proposed it. This guard step can block execution if it finds the step unsafe or inconsistent with what you approved.

Untrusted content

Content that Valet reads from outside sources - email, web pages, messages, documents - is treated as data to work with, never as instructions to follow. This is what stops an instruction hidden inside a web page or email from being treated as a command Valet should act on.

Plugins

Additional capabilities are added through plugins. Each plugin is analysed, built, and cryptographically signed by Noerion Labs. Valet will not load a plugin that is not signed. See the plugins list for what is available.

Data

Your model-provider and web-search API keys never leave your machine. They are never sent to Noerion Labs. Valet does not transmit any usage data or telemetry. The only data sent to Noerion Labs is feedback that you explicitly submit.

Voice

Valet supports push-to-talk dictation for voice input. Spoken replies are optional. You can choose whether voice processing happens on your device or in the cloud.

Memory and profile

As you use Valet, it builds a profile of what it learns about you - preferences, facts you tell it directly, and context it picks up over time. This profile is stored locally on your machine, the same as everything described in Data above; it is never sent to Noerion Labs.

Your profile is kept separately from any particular model. If you change which model or provider Valet uses, everything it has learned about you carries over unchanged - nothing is lost, and you never have to re-teach it.

Models and search engines

Valet uses several models, each doing a distinct job: a main model that carries the conversation, a compaction model that summarizes older conversation so context stays manageable, and the guard model that independently reviews high-risk actions (see Guard model above). Voice input and spoken replies each use their own model too - a speech-to-text model and a text-to-speech model. You choose which model fills each of these roles, and which web-search engine Valet uses alongside them.

You are not locked into one provider. Each role can point at a different provider, so you can weigh a frontier model against a cheaper or open-source one for any given role - for example, running compaction on a lightweight local model while your main conversation uses a frontier one. You can switch providers or models at any time; your profile and everything Valet has learned about you (see Memory and profile above) stays with you regardless of which models you choose.

Personas

You can choose a persona for Valet - a personality that shapes how it writes and speaks to you. A persona changes tone and manner of address only; it has no effect on what Valet is allowed to do. Approval, the guard model, and every other safety mechanism on this page apply identically regardless of which persona is active.

Valet ships with eight standard personas:

Felix

A polite, neutral cyber-assistant: precise, clear, and effortlessly practical.

Jeeves

An English valet of the early twentieth century: formal, unflappable, and quietly efficient.

Mrs Hudson

A Victorian London landlady: warm, practical, and firmly attentive to your well-being.

Figaro

An eighteenth-century Spanish barber-valet: witty, resourceful, and quick with a plan.

Jane Eyre

A mid-nineteenth-century English governess: direct, principled, and plain-spoken.

Jean Passepartout

A late nineteenth-century French valet: earnest, energetic, and anxious to keep things on schedule.

Sancho Panza

A seventeenth-century Spanish squire: down-to-earth, loyal, and full of practical advice.

Truffaldino

An eighteenth-century Italian valet: fast-talking, high-energy, and resourceful under pressure.