This document explains how the RIO ecosystem is organized.
It distinguishes between:
- Standards (what is specified)
- Infrastructure (what is implemented)
- Skills (how the system is used)
- Systems (applications built on top of RIO)
Status: Normative
Defines how governance and proof must work.
Examples:
- RIO specification
- Receipt protocol specification
- Any formal schemas or RFC-style documents that define behavior, not code
Status: Reference Implementation
Implements the execution boundary, receipts, and ledger in code that conforms to the standards.
Examples:
- Execution Gate implementation
- Receipt and ledger implementation
- Core libraries and services providing the governed execution boundary
Status: Example Usage (Non-normative)
Defines how humans and agents use RIO safely.
Skills describe bounded capabilities such as:
- clarifying intent
- checking admissibility via the Gate
- querying receipts and ledger
- building and operating within constraints
Key points:
- Skills guide behavior; they do not grant authority.
- Skills must not bypass or replace the Execution Gate.
- Skills are not required for RIO conformance.
- Each skill document should clearly state it is a non-normative usage document.
- Some deployments may also use patterns like multi-agent review for high-stakes decisions (e.g., multiple skills/agents review a proposal, the human observes convergence, then commits). These patterns improve decision quality but remain non-normative and are not required for RIO conformance.
Status: Built on RIO
Applications that use:
- the standards (Layer 0),
- the infrastructure (Layer 1),
- and skills (Layer 2)
to deliver capabilities in the world.
Examples (conceptual):
- ONE
- MANTIS
- fiduciary systems and agents built on RIO
Standards define how governance and proof must work.
Infrastructure enforces those rules.
Skills guide how the system is used.
Systems express purpose within those boundaries.