Policy Examples

Retail refund policy

• Allow refund.create under USD 250 when the order is refundable and no risk flag is present.
• Require approval from USD 250 to USD 500.
• Deny above USD 500, duplicate refunds, stale refund windows, suspicious reason codes, and three or more same-customer attempts in 24 hours.

Customer data policy

• Deny unnecessary customer.pii.export from autonomous agents.
• Require human approval for sensitive reads tied to support investigations.
• Redact payload evidence before it reaches public surfaces.

Production write policy

• Deny destructive writes outside maintenance windows.
• Require approval for schema changes, bulk updates, and privileged operations.
• Quarantine agents that repeat denied write attempts.

Implementation notes

Keep the HaltState call as close as possible to the side effect. The agent may plan and draft freely, but the wrapper around the actual action should be the place where authority is checked. That wrapper should send only the context required for policy evaluation: safe identifiers, normalized amounts, action names, risk flags, schedule windows, and redaction status. Raw customer payloads and secrets should stay in the business system or protected operator tooling.

Operational evidence

For each action, preserve the decision, the worker outcome, the idempotency key, safe resource references, latency, proof status, and redaction status. This evidence supports incident response and control narratives because it shows what the system did at runtime rather than only describing what the policy document intended. HaltState supports alignment work; it is not a substitute for legal advice or a compliance certification.