Provenance and Reconstructability
Research question
Can what a system did be reconstructed later, by someone who was not there?
Working thesis
Determinism, retained traces, and public artifact registers are what make a claim checkable after the fact. Reproducibility, traceability, and recovery are properties designed into the substrate — they are not recovered heroically after something goes wrong.
A thesis is the lab’s current position, not a result. What supports it is listed below; what remains unresolved is stated at the end of this page.
Publications
4 attributedDeterminism Is All You Need
Toward Replayable, Governed, and Transactional AI Systems
Research ReleasePublic DraftMarch 17, 2026Evidence-Led Agent Workflows
Agents should produce verifiable artifacts as they work.
Architecture NotePublishedOctober 3, 2025Controlled Extraction from Flagship Systems
Public components should be extracted, not exposed.
Architecture NotePublishedSeptember 3, 2025Public Artifact Registers as Trust Infrastructure
A list of what exists is a form of governance.
Architecture NotePublishedMay 17, 2025
Apparatus
2 systemsSystems attached to this program are what the lab builds to investigate the question. Maturity, evidence level, and current activity are recorded separately on each system — a system attached here is not thereby claimed to be production-ready.
StrongHold
Adaptive Storage & Archive Intelligence
A smart storage system for adaptive compression, archival intelligence, retrieval, and restoration — storage that understands artifact structure and can explain its own strategy decisions.
Artchiv
Creative Media Archive & Provenance Platform
The source of truth for creative media — a governed archive for versioned assets, provenance records, retrieval, and inspectable media lineage across studios, agencies, and production workflows.
Open questions
Published as open. These are the parts of the program the lab does not consider settled.
- 01
How much can be retained before retention becomes its own liability?
- 02
What is the minimum record that makes a run genuinely reconstructable?
- 03
How is provenance kept intact across a boundary the lab does not control?