Every Action Is Visible
Promise
You always know what the AI is doing. It acts in places you can inspect.
How we do it
Each operation has a clear path and visible status, from preparation to execution to completion.
Trust Center
Security and privacy you can actually see and control.
Pause it, adjust it, or trace it at any time.
Lynx keeps perception and action inside clear boundaries: permissioned, auditable, and interruptible. Trust comes from the system itself, not from slogans.
Commitments & Controls
You Decide How Far It Goes
Promise
You decide how far the AI goes. Each task can run at a different level of automation.
How we do it
Lynx supports different execution modes: analysis only, suggested action, or fully automated execution. You can choose the right level for each task.
You Can Stop Anytime
Promise
AI remains interruptible. You can pause, cancel, or adjust while it works.
How we do it
Tasks can be stopped during execution, key operations ask for confirmation, and completed steps remain traceable.
Only The Needed Scope
Promise
It sees the context required for the task and avoids unrelated data access.
How we do it
Context collection is scoped to the current task, without broad scans or unrelated reading.
Stronger Ability, Clearer Boundaries
Promise
As capability grows, control stays explicit. AI actions remain permissioned.
How we do it
From perception to execution, permissions cover external messages, file operations, and cross-app behavior.
Trust & Privacy, Built In.
Your Data Stays Yours.
Sovereignty1.
1.
Local By Default
Lynx runs on your machine. Files stay on-device; system context stays local—no shadow cloud sync and no silent harvesting.
Isolation2.
2.
Sandbox First
Work executes inside controlled boundaries with sandboxing wherever possible so capability stays fenced away from your core system.
Consent3.
3.
Explicit Permissions
Lynx touches only what you approve, respects macOS privacy surfaces, and keeps scopes visible end to end.
Confidence4.
4.
Trust In Practice
Built so daily use—mail, docs, reminders, creative files—doesn't come with second-guessing where data lands.
Every Step Has Boundaries, Logs, and Controls.
Some Things It Will Not Do
Lynx works only within the scope you authorize. Data access and external actions require explicit permission.
All Operations Are Auditable
Each step is recorded and traceable, so you can see what happened and how the system made its decision.
AI Can Be Turned Off Anytime
Lynx is an enhancement layer. You can turn it off at any time, and your apps, data, and workflow stay exactly as they are.
Default Minimization And Restraint
The system collects only necessary context for the current task and prioritizes suggestions before action.
