Design · Governance · Execution · Trust
Build AI systems that know who they are, why they act, and how to stay accountable.
Modern AI systems are powerful — but unstructured.
They lack stable identities, explicit roles, governance boundaries, portable behavior definitions, and shared standards across tools and platforms.
The Persona Management Layer organization exists to solve this problem.
We design open standards, protocols, and reference implementations that make personas a first-class primitive in AI systems.
Personas are not UI features.
They are architectural control units.
A persona defines intent, scope, behavioral constraints, output contracts, and governance rules.
- ARAL — Agent Role Abstraction Layer
- PML — Persona Management Layer
- PCL — Persona Control Language
- ADP — Agent Delegation Protocol
- Determinism First
- Persona Before Prompt
- Governance by Design
- Portability
- Auditability
- Security by Construction
- Human-Centered AI
Standards: Draft → Community Review
Implementations: Experimental
Welcome to the layer AI was missing.