Intelligence vs Authority
Intelligence and authority are not the same.
Many autonomous systems blur this distinction. If a system can reason, it is allowed to act. If it understands, it executes.
ClawPaw intentionally separates cognition from authority.

Autonomy modes exist because understanding does not imply permission. Planner Mode exists because reasoning should be visible before execution. Structured decision interfaces exist because ambiguity should not default to action.
The system can generate a complete execution plan without being authorized to carry it out.
Structural Separation
ClawPaw enforces boundaries at multiple layers:
- Autonomy Mode determines action scope
- Permission scope limits system access
- Folder boundaries restrict file operations
- macOS privacy controls remain authoritative
Authority is layered and adjustable.
Why It Matters
Powerful systems become fragile when cognition and authority collapse into one layer.
ClawPaw is designed to remain intelligent without becoming uncontrolled.
Understanding informs action. It does not mandate it.
