Standardized Modular Contextual Protocols

AI Governance. Every Agent Authenticated. Every Action Logged.

AI Is Already Running in Your Practice. Do You Always Know What It’s Doing?

Standard AI protocols let agents communicate with systems. SMCP adds what healthcare specifically requires: authentication for every agent (no anonymous access), least-privilege data access, human approval gates that physically block payer submissions until your team approves, full immutable audit logging of every interaction, and cost attribution per transaction.

SMCP is a standalone product — used internally by AROS and licensable to any healthcare organisation running AI agents on clinical and billing data.

Every AI agent authenticated and credentialed before accessing any practice data
Least-privilege access — field-level permissions per agent template, not broad system access
Human approval gates: ADR responses and denial appeals NEVER auto-submit — hard constraint, cannot be disabled
Immutable audit log — agent ID, data accessed, action taken, timestamp, user, outcome — 7-year retention
Cost attribution per transaction — every Availity call and LLM inference logged to the specific claim
Real-time cost transparency dashboard — no surprise bills
Standalone licensable product — deploy for any AI running on healthcare data