SMALL
The State Protocol for Agent-Legible Systems
AI agents don't fail because models are weak.
They fail because the systems around them are implicit, mutable, and opaque.
SMALL replaces guesswork with execution contracts.
Schema → Manifest → Artifact → Lineage → Lifecycle
SMALL is a state protocol that makes systems legible to machines. It defines how intent is declared, how state is created, how changes are traced, and how lifecycle is governed - deterministically.
SMALL governs state, not execution.
It makes implicit systems explicit, mutable systems auditable, and opaque systems governable.
- Schema - what is allowed
- Manifest - what is intended
- Artifact - what exists
- Lineage - where it came from
- Lifecycle - what happened and what may happen next
If a system cannot answer these questions explicitly, it is unsafe for agents.
Getting Started
New to SMALL? Start here.
Learn who writes what, how the CLI works, and get a copy-paste starter prompt for your agent.
SMALL already runs.
It enforces failure, rollback, and auditability by design.
Start with the CLI to scaffold and validate SMALL artifacts in your project.