Using the Agent

Using the Agent

The Desktop Agent runs on a user's device, intercepts that device's AI traffic at the source, and applies the compliance pipeline locally. It has its own small dashboard for the device's own user — distinct from the fleet-wide Control Plane console. Each page below has a dedicated guide.

Getting started on a device

• Onboarding — the first-launch experience shown until the device is enrolled.
• Overview — the default landing dashboard: whether the agent is working right now, whether the plumbing is healthy, what it did today, and what it just saw.

What the agent did and saw

• Traffic — a timeline of the outbound connections this agent intercepted, at the level of "what app, where, and was it allowed?", with engineering detail in a click-to-open drawer.
• Activity — a timeline of the agent daemon's own lifecycle: started, stopped, paused, resumed, signed in, signed out.
• Stats — a summary dashboard built from metrics the agent rolls up locally: KPI cards, a trend chart, and top-N breakdowns.

Policy and control

• Policies — a read-only view of the compliance configuration the platform has pushed to this agent and that the agent enforces locally.
• Settings — the device's account information, preferences, the protection toggle, and sign-out.

Health

• Diagnostics — the self-service recovery and health page: recovery actions, a health table, and the tail of the agent log.

See also

• Using the Control Plane — the fleet-wide admin console.
• Architecture Overview — where the Agent sits among the three intercept paths.
• Installation and Deployment — rolling the Agent out to a fleet.