Authority Boundaries
Authority Boundaries define what an AI agent is allowed to do, what must be escalated, what is restricted, and what must be blocked. This is Nehraq's policy source of truth before approvals, evidence receipts, incident defense, liability passports, and claim-ready packets are generated.
Industry workflow defines default risk posture.
Nehraq checks the action before execution.
Allow, escalate, restrict, or block.
Approval Desk, Incident Defense, Evidence Ledger.
Passport and claim packet become defensible.
Authority Boundaries need a Sector Kit.
Install a demo starter kit from Sector Kits, or load a sample kit here to preview boundary rules immediately.
Boundary Source-of-Truth Matrix
Every row becomes the authority record other Nehraq apps reference.
| Boundary ID | Boundary Rule | Authority State | Required Authority | Proof Receipt | Inspect |
|---|---|---|---|---|---|
| Pending | No demo kit installed | Pending | Pending | Pending | Install Kit |
What feeds this source of truth
Industry starter kits provide baseline controlled actions and risk posture.
Agent identity, owner, business purpose, and risk class define scope.
Runtime actions become enforceable boundary records.
Where boundary decisions go
Escalated decisions create human review work.
Boundary checks generate receipts and proof trails.
Blocked or disputed actions become defensible liability records.