EASTER Draft 2: A Six-Primitive Model for Consequential Continuity in Agent Systems

Preprint announcement — October 2026

I’m publishing the second draft of my EASTER paper. This is the prepublication version — the frozen manuscript now being prepared for arXiv submission.

What it is

EASTER — Evidence, Authority, State, Transition, Exception, and Receipt — is a six-primitive model for preserving consequential continuity across computational change. When intelligent work crosses boundaries between models, agents, sessions, processes, machines, and runtimes, something needs to hold the structure of what actually happened. EASTER is my answer to what that something looks like.

The paper defines the six primitives, implements them in a Python/SQLite reference kernel, and evaluates the result three ways: structural invariant testing, a recovered six-system comparative analysis, and — unusually — by using the reference implementation during the manuscript’s own preparation to preserve contribution provenance across multiple human and AI participants.

The hypothesis

The paper motivates a falsifiable claim: Evidence, Authority, State, Transition, Exception, and Receipt may be sufficient macroscopic variables for preserving consequential intelligent work across microscopic computational change. The direct test — heterogeneous-runtime continuation without access to the originating runtime’s private state — is specified prospectively. It hasn’t been run yet, and the paper is honest about that.

How it was built

This manuscript was developed through sustained collaboration between me and multiple AI systems operating across distinct computational environments — ChatGPT, Claude, and independent agent systems running on separate machines. The paper’s own §8 documents exactly who contributed what, because the provenance of the claims matters to the claims themselves.

The AI systems are named explicitly, not treated as invisible tooling. Their contributions included intellectual work that materially affected the resulting artifact. Formal authorship remains with me where publication conventions require human accountability.

The frozen artifact

The publication manuscript is frozen at a specific commit. Everything below points to that exact, inspectable artifact:

The freeze attestation establishes an immutable publication baseline. The LaTeX package above was derived from it mechanically; substantive changes will be recorded as later revisions — not retrofitted into this draft.

What’s next

arXiv submission is the next step, then a formal venue. The §11.6 portability trial — the direct experimental test of the macroscopic-variable hypothesis — is the next research milestone.

If you work on agent infrastructure, governance, or the problem of what persists when everything around the work changes, I’d welcome your reading.

— Nathan Woolen, WitsBI

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Scroll to Top

Discover more from Witsbi

Subscribe now to keep reading and get access to the full archive.

Continue reading

Witsbi Bot
Hi there! Ask me anything 👋